German Phonetics and Phonology: Theory and Practice 9780300225181

Citation preview

German Phonetics and Phonology

This page intentionally left blank

German Phonetics and Phonology

Theory and Practice

mary grantham o’brien  sarah m. b. fagan

  New Haven & London

Copyright © 2016 by Yale University. All rights reserved. This book may not be reproduced, in whole or in part, including illustrations, in any form (beyond that copying permitted by Sections 107 and 108 of the U.S. Copyright Law and except by reviewers for the public press), without written permission from the publishers. Yale University Press books may be purchased in quantity for educational, business, or promotional use. For information, please e-mail [email protected] (U.S. office) or [email protected] (U.K. office). Editor: Sarah Miller Publishing Assistant: Ashley E. Lago Manuscript Editor: David Pritchard Production Editor: Ann-Marie Imbornoni Production Controller: Maureen Noonan Designed by Newgen North America. Set in Gentium Plus and Minion Pro. Printed in the United States of America. Library of Congress Control Number: 2016941499 ISBN: 978-0-300-19650-4 (paperback : alk. paper) A catalogue record for this book is available from the British Library. This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). 10 9 8 7 6 5 4 3 2 1

Contents

List of Figures xi List of Tables xiii Preface xv Acknowledgments xvii

Chapter 1. Phonetics 1 1.1 . Introduction 1 1.2 . Pronunciation Standards 2 1.3 . Phonetic Transcription 3 1.4 . The Organs of Speech 4 1.5 . Classifying Sounds 7 1 .5.1. Classifying Consonants 8 1 .5.2. Classifying Vowels 10 1.6 . The Phones of German 13 1 .6.1. The Consonants 13 1 .6.2. The Vowels 17 1.7 . Using IPA Symbols 19 1.8 . Advanced Topics 20 1 .8.1. Consonants 20 1 .8.2. Vowels 21 Exercises 22

Chapter 2. Acoustic Phonetics 29 2.1 . Introduction 29 2.2 . Waveforms and Spectrograms 31 2.3 . The Acoustics of German Vowels 34 2.4 . The Acoustics of German Consonants 35 2 .4.1. The Acoustics of German Obstruents 36 2 .4.2. The Acoustics of German Sonorants 37 2.5 . Advanced Topics 38 Exercises 40

vi  Contents

Chapter 3. Phonology 43 3.1. Introduction 43 3.2. Basic Terms 43 3.3. The Consonant Phonemes of German 46 3.3.1. Phonemic Contrasts 47 3.3.2. Allophonic Variation 48 3.4. The Vowel Phonemes of German 51 3.4.1. Phonemic Contrasts 51 3.4.2. Allophonic Variation 54 3.5. Advanced Topics 55 3.5.1. Nasalized Vowels 55 3.5.2. The Allophones of /ʀ/ 56 Exercises 57

Chapter 4. Syllables and Word Stress 63 4.1. Introduction 63 4.2. Sonority 63 4.3. Phonotactics 64 4.4. Building Syllables 67 4.5. Word Stress 75 4.5.1. Simplex Words 76 4.5.2. Complex Words 76 4.5.2.1. Suffixes 77 4.5.2.2. Prefixes 78 4.5.3. Compound Words 80 4.6. Advanced Topics 81 4.6.1. Ambisyllabic Consonants 81 4.6.2. Syllable Boundaries and Word Boundaries 82 4.6.3. Stress in Compounds 84 Exercises 85

Chapter 5. Prosody 93 5.1. Introduction 93 5.2. Rhythm 94 5.3. Phrasal Stress 96 5.4. Sentence Stress and Focus 97 5.4.1. Broad Focus 98 5.4.2. Narrow Focus 99 5.5. Intonation 101 5.5.1. Intonation Phrases 101

Contents  vii 5.5.2. Multiple Intonation Phrases 102 5.5.3. Intonation Contours 104 5.6 . Advanced Topics 105 Exercises 107

Chapter 6. Alternations 113 6.1 . Introduction 113 6.2 . Automatic Alternations 114 6 .2.1. Vowel Shortening 114 6 .2.2. Final Devoicing 115 6 .2.3. Dorsal Fricative Assimilation 115 6 .2.4. Spirantization 117 6 .2.5. Glottal Stop Insertion 118 6 .2.6. Schwa Deletion 119 6 .2.7. R-Vocalization 122 6 .2.8. Sonorant Syllabification 122 6 .2.9. Nasal Assimilation 123 6.3 . Morphophonemic Alternations 124 6 .3.1. Ablaut 124 6 .3.2. Umlaut 126 6.3.2.1. Noun Plurals 128 6.3.2.2. Comparative and Superlative 128 6.3.2.3. Present Tense of Strong Verbs 129 6.3.2.4. Subjunctive II 129 6.3.2.5. Derivation 130 6.4 . Advanced Topics 131 6.4.1. Voice Versus Spread Glottis 131 6.4.2. Glottal Stop 133 Exercises 134

Chapter 7. Historical Developments and Regional Variation 141 7.1 . Introduction 141 7.2 . Types of Sound Change 141 7.3 . Sound Changes in the History of German 143 7 .3.1. Old High German 145 7 .3.2. Middle High German 147 7 .3.3. Early New High German 149 7 .3.4. Summary 150 7.4 . Development of German Spelling 152 7.5 . Regional Variation 157

viii  Contents 7.5.1. Colloquial German 157 7.5.2. German Dialects 159 7.6. Advanced Topics 163 7.6.1. Germanic Accent Shift 163 7.6.2. Verner’s Law 163 Exercises 165

Chapter 8. Phonetics in the Real World 169 8.1. Introduction 169 8.2. Orthography and the IPA: Letters and Sounds 169 8.2.1. Consonants 169 8.2.2. Vowels 177 8.2.3. Advanced Topics 182 8.3. Reading the IPA 183 8.4. Transcription I: From the Written Word 185 8.5. Pronunciation Similarities and Differences 191 8.6. Pronunciation Practice I 192 8.6.1. New German Consonants 193 8.6.1.1. Perceptual Practice 1 195 8.6.1.2. Production Practice 1 198 8.6.2. New German Vowels 200 8.6.2.1. Perceptual Practice 2 202 8.6.2.2. Production Practice 2 206

Chapter 9. Acoustic Phonetics in the Real World 209 9.1. 9.2. 9.3. 9.4. 9.5.

Introduction 209 Making Use of Computer Technology 209 Speech Analysis I: Waveforms and Spectrograms 210 Analyzing German Vowels 213 Analyzing German Consonants 217

Chapter 10. Phonology in the Real World 219 10.1. Introduction 219 10.2. Minimal Pairs 220 10.3. Transcription II: From Grapheme to Phoneme 223 10.3.1. Consonants 224 10.3.2. Vowels 230 10.4. Pronunciation Practice II 235 10.4.1. Similar German Consonants 236

Contents  ix 10.4.1.1. Perceptual Practice 1 238 10.4.1.2. Production Practice 1 241 10.4.2. Similar German Vowels 243 10.4.2.1. Perceptual Practice 2 245 10.4.2.2. Production Practice 2 247 10.5. Acoustic Analysis: Similarities between German and English 249 10.5.1. Analyzing Similar Consonants 249 10.5.2. Analyzing Similar Vowels 252

Chapter 11. Working with Syllables and Word Stress 257 11.1. Introduction 257 11.2. Syllabification 258 11.3. Determining Word Stress 267 11.3.1. Default Word Stress 268 11.3.2. Affixes and Word Stress 269 11.3.2.1. Suffixes 270 11.3.2.2. Prefixes 272 11.3.3. Stress in Compound Words 275 11.4. Acoustic Analysis: Correlates of Word Stress 278

Chapter 12. Making Use of Prosody 281 12.1. Introduction 281 12.2. Perceiving and Producing Rhythm 282 12.3. Working with Phrasal Stress 284 12.4. Determining Sentence Stress and Focus 285 12.4.1. Broad Focus 286 12.4.2. Narrow Focus 287 12.5. Intonation 289 12.5.1. Intonation Contours in Statements and Questions 292 12.5.2. Other Intonation Contours 293 12.6. Acoustic Analysis: Sentence Stress and Intonation Contours 299 12.6.1. Analyzing Sentence Stress 299 12.6.2. Analyzing Intonation Contours 300 12.6.3. Comparing German and English Intonation Contours 301

Chapter 13. Alternations in the Real World 303 13.1. Introduction 303 13.2. Vowels 304 13.2.1. Vowel Shortening 304

x  Contents 13.2.2. Vowel Centralization 305 13.2.3. Vowel Reduction 306 13.2.4. Schwa Deletion 307 13.3. Consonants 313 13.3.1. Assimilation 313 13.3.2. Deletion 316 13.4. Weak Forms 319 13.5. Acoustic Analysis: Alternations 323 13.5.1. Vocalic Alternations 323 13.5.1.1. Vowel Shortening 324 13.5.1.2. R-Vocalization 325 13.5.2. Consonantal Alternation: Spirantization 325

Chapter 14. Varieties of German 327 14.1. Introduction 327 14.2. Sound Change 328 14.2.1. Cognates 329 14.2.2. German Surnames in English 332 14.2.3. Sound Changes Today: North American English 334 14.3. Modern German Dialects 337 14.3.1. Segmental Differences 337 14.3.2. Prosodic Differences 341 14.4. Acoustic Analysis 342 14.4.1. Final Devoicing 343 14.4.2. Rounded Vowels 344 14.4.3. Affricates 345 Appendix A: The Sounds of German 347 Appendix B: IPA Symbols 349 Appendix C: Affixes and Stress 351 Appendix D: Pronunciation of Vowel Letters 355 Glossary 357 Data Sources 371 References 373 Index 379

Figures

 1.1  1.2  1.3  1.4  2.1  2.2  2.3  2.4  2.5  2.6  2.7  2.8  5.1  7.1  7.2  7.3  7.4  9.1  9.2  9.3  9.4 10.1 10.2 10.3 10.4 10.5 11.1

The Organs of Speech 4 The Larynx 5 The Tongue 7 The IPA Quadrilateral for Vowels in German 10 Waveform of a Periodic Signal 32 Waveform and Spectrogram of Human Speech 33 Spectrogram of German Vowels with Formants Marked 35 Spectrogram of German and  35 Spectrogram of German  36 Turbulent Noise in German Fricatives 37 Spectrogram of German  37 Voice Onset Time in German  39 British School Transcription of  106 The Major Modern German Dialect Areas 145 Primary Umlaut 147 Secondary Umlaut 148 Isoglosses of the Second Sound Shift 160 Waveform of German  212 Spectrogram of German  213 Spectrogram of German with Vowel Steady States Outlined 214 Spectrogram of German  217 [l] in German and [ɫ] in American English  250 German and English  251 German  252 German and American English  253 German and  254 German  279 xi

xii  List of Figures 12.1 12.2 12.3 12.4 12.5 14.1 14.2

Intonation Contours of the Declarative Utterance and the Yes-No Question  290 Spectrogram and Intonation Contour of the Utterance  291 Spectrogram and Intonation Contour of the Utterance  291 Intonation Contour of the Question  292 Spectrogram and Intonation Contour of the Utterance  294 Spectrogram of German and English  343 Spectrogram of German  345

Tables

 1.1  1.2  2.1  3.1  3.2  4.1  4.2  6.1  6.2  6.3  7.1  7.2  7.3  7.4  7.5  8.1  8.2  8.3  9.1 10.1 10.2 13.1 13.2 13.3 13.4 14.1

The Consonant Sounds of German 14 The Vowel Sounds of German 17 Mean Formant Values for German Vowels 40 The Consonant Phonemes of German 46 The Vowel Phonemes of German 52 Consonant Clusters at the Beginning of German Syllables 66 Consonant Clusters at the End of German Syllables 66 Restrictions on Schwa Deletion 120 Ablaut Patterns 125 Umlaut Alternations 127 The First Sound Shift 144 The Second Sound Shift 146 Sound Changes in the History of German 151 Sound Changes in the Modern German Dialects 162 Alternations in Modern Standard German Resulting from Verner’s Law 164 Sound-Grapheme Correspondences for German Consonants 177 Grapheme-Sound Correspondences for German Monophthongs 181 Grapheme-Sound Correspondences for German Diphthongs 182 German Vowel Formant Values 215 Grapheme-Phoneme Correspondences for German Consonants 225 Grapheme-Phoneme Correspondences for German Vowels 233 Weak Forms: Pronouns 321 Weak Forms: Articles and Relative/Demonstrative Pronouns 321 Weak Forms: Auxiliary and Modal Verbs 322 Weak Forms: Prepositions, Conjunctions, and Adverbs 322 Consonant Correspondences in German and English Resulting from the Second Sound Shift 330 xiii

xiv  List of Tables 14.2 14.3 14.4 14.5 A.1 A.2 B.1 B.2 C.1 C.2 C.3 C.4 D.1 D.2

Additional Consonant Correspondences in German and English 331 The Second Sound Shift in the Modern German Dialects 338 Diphthongization in the Modern German Dialects 339 Monophthongization in the Modern German Dialects 340 The Consonants of German 347 The Vowels of German 348 IPA Symbols for Vowels and Consonants Used in This Book 349 IPA Diacritics Used in This Book 350 Suffixes That Do Not Bear Primary Stress 351 Suffixes That Bear Primary Stress 352 Unstressed Prefixes 352 Prefixes That Can Bear Primary Stress 353 Short Vowels Before One Consonant Letter in the Same Morpheme 355 Long Vowels Before Two or More Consonant Letters in the Same Morpheme 355

Preface

This book combines theory and practice to provide students with an introduction to the phonetics and phonology of Modern German and the means to apply this knowledge to improve their pronunciation of the language. The theory discussions do not focus on linguistic argumentation, nor do they provide exhaustive proof for the conclusions that are reached; theory is used as a tool. The ultimate goal of this book is to help students make their speech more comprehensible. It is intended as a textbook for North American undergraduates who are in various stages of their German careers. While a basic knowledge of German is expected, no background in linguistics is necessary. Linguistic terminology is explained as it is introduced and a glossary of linguistic and technical terms is also provided for easy reference. The book is divided into two sections, theory and practice, and each chapter in the first section has a corresponding chapter in the second section. Chapter 1 provides an introduction to the field of phonetics and presents a brief description of the articulatory characteristics of the speech sounds of German. This chapter provides the foundation for the remaining chapters in the book and should be worked through carefully before proceeding to other chapters. Chapter 2 is a basic introduction to acoustic phonetics (the study of the physical properties of speech sounds). Chapter 3 introduces the key concepts of phonology (the study of sound systems) and presents the phonemes of German, the distinctive sounds of the language. Syllable structure and word stress are treated in chapter  4; rhythm, sentence stress, and intonation are presented in chapter 5. Chapter 6 discusses the common segmental alternations found in German, alternations like final devoicing and umlaut. The last chapter in the theory section, chapter 7, deals with historical developments and regional variation, focusing on the sound shifts that have shaped the language over time and the phonetic differences that distinguish speakers from the major regions of the German-speaking area. Each chapter in the theory half of the book includes a section that addresses advanced topics. Exercises at the end of each chapter provide students with the opportunity to work with and apply the information presented. The seven companion chapters in the practice section (chapters 8 through 14) provide exercises and activities that help students put theory into practice. They include exercises that focus on perception and production—not just of ­individual xv

xvi  Preface speech sounds, but also of aspects of speech that that go beyond individual sounds (word stress, sentence stress, intonation). Numerous exercises are provided for practice in phonetic transcription (these can be found in the theory portion of the book as well). Chapter 9 familiarizes students with tools they can use to analyze the acoustic properties of speech. Each subsequent chapter in the practice half of the book contains an acoustic analysis section that provides students with the opportunity to analyze the acoustic properties of their German speech—the properties of individual sounds and aspects of speech at the word and sentence level. The goal of the activities in these sections is to show students how their speech differs from that of native speakers and to provide them with strategies for making their German more comprehensible. It is possible to make use of this book in a number of ways. Chapters in the first half can be used by themselves in German phonetics and phonology courses that focus primarily on theory. Chapters in the second half of the book lend themselves well to courses that focus on developing German students’ speaking and listening skills. It is also possible to make use of the entire book in courses that focus on developing students’ theoretical and practical skills. In this instance we recommend working with corresponding pairs of chapters (chapters 1 and 8 followed by chapters 2 and 9, and so on). Regardless of the way in which you make use of the book, we recommend working through chapter 1 as a first step. The Web site that accompanies this book (yalebooks.com/germanphonetics) provides the sound files for the exercises in the second half of the book. It also includes recordings of the sounds of German and sample words for each sound, found in appendix A. An answer key for the exercises in the book is available to instructors on a separate section of the Web site.

Acknowledgments

We are very grateful to our colleagues who have contributed to this project. We owe special thanks to John te Velde, who provided feedback on the entire manuscript, and to Michael Jessen, who read many of the chapters with a critical eye. A number of other colleagues provided detailed comments and valuable suggestions for the improvement of individual chapters. These include Tracy Alan Hall, Carrie Jackson, Mark Louden, Ines Martin, Richard Page, Ashley Roccamo, Dorian Roehrs, and Christopher Sapp. Special thanks to Ingo Titze for permission to use modified versions of images from Principles of Voice Production. We would also like to thank Peter Auer, Johannes Knaus, Bettina Knaus, and Joseph Salmons for their assistance with recordings, and Jamie O’Brien for his help with figures. Thanks also go to our students at the Universities of Iowa and Calgary who provided us with practical feedback on the effectiveness of exercises and on the clarity of our explanations.

xvii

This page intentionally left blank

German Phonetics and Phonology

To access the German Phonetics and Phonology Web site, visit yalebooks.com/germanphonetics.

Chapter 1

Phonetics

1.1. Introduction You may not have given much thought to the way in which speech sounds are produced until you started learning to speak German. When you encountered words like schön ‘beautiful’, müde ‘tired’, and rot ‘red’, which contain sounds that do not exist in English, did you wonder what you had to do with your tongue, lips, and other parts of your mouth to make those new sounds? Did this lead you to wonder how speech sounds are produced in general? Did you also think about the range of sounds that are used in the languages of the world? These kinds of questions are of interest to people who study phonetics, the branch of linguistics that deals with the sounds of human speech. Regardless of the language being spoken, speech is an act of communication that involves a speaker and a hearer. The speaker produces speech sounds, which are transmitted to the hearer, who receives and interprets them. These three phases of the communication process—production, transmission, and reception of speech sounds—are the focus of the three main subfields of phonetics: articulatory, acoustic, and auditory phonetics. Articulatory phonetics deals with how people produce speech sounds. This subfield of phonetics is particularly important to language learners because an understanding of the ways in which the different sounds of a language are produced can help students learn to articulate these sounds themselves. The discussion of phonetics in this chapter focuses on articulatory phonetics. Acoustic phonetics is the branch of phonetics that examines the physical properties of the sound waves we produce when we speak. We can measure properties such as the loudness (amplitude), frequency, and quality of the wave. Recent advances in computer hardware and software make it possible to do most recordings and phonetic analyses on a laptop computer. Chapter 2 provides a discussion of acoustic phonetics, and practical exercises in the second half of the book beginning with chapter 9 enable you to record and analyze your own German speech. Auditory phonetics deals primarily with how hearers perceive speech. A sound wave that successfully makes it to the hearer sets the eardrum in motion. The small bones of the inner ear stimulate endings of the auditory nerve. Once the 1

2  Phonetics brain receives the impulses, the sound wave is sensed as sound. Combinations of speech sounds are often then identified as speech. Listeners are rarely tuned in to the individual speech sounds when they listen to the speech of others but are instead more focused on making meaning out of utterances. When we think about what makes up speech, individual sounds (segments) may come to mind—the l sound in Milch ‘milk’, the o sound in Sonne ‘sun’, and so on. While individual sounds do play a major role in phonetic inquiry, the field of phonetics does not stop its investigations there. We are also interested in those features of speech that occur above the level of the segment. Word stress, rhythm, sentence stress, and intonation, which are all physically present in the speech stream, also belong to the field of phonetics. Chapters 4, 5, 11, and 12 treat these suprasegmental aspects of speech. 1.2. Pronunciation Standards The German language described in textbooks and handbooks is Modern Standard German. It is important to remember that this is a variety that is prescribed as the standard. It is a description of what the official language is supposed to look like. The standardization of the German language has taken place over many centuries. In the fourteenth century, Kanzleisprachen (chancery languages) often served as regional standards in the German-speaking territories. With the advent of the printing press at the end of the fifteenth century, five prominent regional Schreibsprachen (written languages) had emerged, shaped by the chanceries and printers, each showing dialectal differences in pronunciation. This path to standardization continued with the translation of the Bible into German in the sixteenth century. Through the efforts of grammarians, lexicographers, and scholars in the seventeenth and eighteenth centuries, who endeavored to shape a written norm, a supraregional standard began to establish itself. Although this eventually led to a standardized written language, it was not until 1898 that a standardized pronunciation was prescribed. It was in this year that Theodor Siebs published his Deutsche Bühnenaussprache, a guide for pronunciation on stage (Siebs 1898). Known as “der Siebs,” this work was long regarded as the authoritative guide for the pronunciation of German. More recently, several additional pronunciation dictionaries have been published. The Duden Aussprachewörterbuch, first published in 1962, established a standard pronunciation (Mangold and Grebe 1962). It was based on research conducted in West Germany. Krech and Stötzer’s Grosses Wörterbuch der deutschen Aussprache (Krech and Stötzer 1982), based on research carried out in East Germany, was published in 1982. Most recently, Krech and colleagues have published Deutsches Aussprachewörterbuch (Krech et al. 2009), which includes information on the standard pronunciation of German in Austria and in the ­German-speaking area of Switzerland.

Phonetics  3 Unless we indicate otherwise, the pronunciation of German presented here is the standard pronunciation (Standardlautung) prescribed in Duden: Aussprachewörterbuch (Mangold 2005), the sixth edition of the Duden pronunciation dictionary. You have no doubt already encountered spoken German that deviates from the standard. For example, you may have heard a speaker pronounce the word Hoffnung ‘hope’ as Hoffnunk (with a k at the end). This is typical of speakers from northern and central Germany. Or you may have noticed someone pronouncing the g in dreißig ‘thirty’ as a k rather than like the ch in the word ich ‘I’. This is typical of speakers from southern Germany. These sorts of pronunciations are not wrong. They are simply examples of other varieties of German. Although we will focus on standard pronunciation, we will also provide information on pronunciation in other varieties of the language. Regional variation in pronunciation will be treated in chapter 7. 1.3. Phonetic Transcription You have probably noticed that German orthography (spelling) provides a good deal of information about pronunciation. For example, when you see a word like Knabe ‘boy’, you know that each letter represents a sound—unlike the word knave in English (related historically to Knabe), which contains two letters that are not pronounced, k and e. In fact, German orthography is relatively transparent in comparison to English. When you see the letter a in a German word, you can be pretty sure that it will be pronounced as an a—either as a long a or a short a (the a in Knabe is long; the a in Ball ‘ball’ is short). In English, on the other hand, the letter a can represent at least six sounds that differ in ways other than their duration. If you say the following words out loud, you will hear six different ways to pronounce the letter a in English: lame, dad, father, fall, sofa, many. In spite of its relative transparency, German orthography is not ideal for representing the sounds of the language in an explicit way. This is because the relationship between the graphemes of German (the letters of the German alphabet) and the sounds they represent is not one-to-one. That is, we cannot say that each German grapheme represents only one sound or that each sound is represented by only one grapheme. The grapheme , for example, can represent at least three different sounds in German, as the following words demonstrate: ‘path’, ‘bed’, ‘sun’. (We put letters in angle brackets, < >, when we want to be explicit that we are talking about orthographic symbols.) The sound represented by the grapheme in the word ‘door’ can also be represented by the grapheme , as we see in the spelling of a word like ‘type’. In order to ensure a one-to-one correspondence between sound and symbol, linguists use special phonetic alphabets to represent speech sounds. We use the International Phonetic Alphabet (IPA), a notational standard for the phonetic

4  Phonetics representation of all languages developed by the International Phonetic Association, the oldest representative organization for phoneticians, established in 1886 in Paris. When we want to be explicit that we are using phonetic symbols, we put them in square brackets, [ ]. To demonstrate the one-to-one relationship between the symbols of the IPA and the sounds of German, we can transcribe the German words we discussed in the preceding paragraph. When we transcribe the words , , and using the IPA, we see that three different phonetic symbols are used to represent the three different sounds in these words: [veːk], [bɛt], [zɔnə]. When we transcribe the words and using the IPA, we see that the same phonetic symbol, [ yː], is used for and : [tyːɐ̯], [tyːp]. 1.4. The Organs of Speech To understand the production of speech sounds, we first need to identify those parts of the human body that are involved in speech production. Take a look at figure 1.1, which illustrates the important organs of speech. Speech sounds are produced when an airstream is set in motion. This is the first phase of speech, initiation. For the sounds of German, the airstream is set in motion by pushing air out of the lungs. This pulmonic egressive airstream, that is, air being pushed out

Figure 1.1 The organs of speech. Modified from I. R. Titze, “Principles of Voice Production: Second Printing.” Copyright 2000 by the National Center for Voice and Speech, Iowa City, IA 52242. Courtesy of Ingo R. Titze.

Phonetics  5 of the lungs, acts as the power source for speech. The air enters the trachea (the windpipe) and then the larynx (the voice box). The next phase of speech, phonation, occurs in the larynx. The larynx is made up of cartilage and muscle, as shown in figure 1.2. The vocal folds (or vocal cords), two pairs of muscular folds, are attached to the inner sides of the thyroid cartilage at the front of the larynx and the two arytenoid cartilages at the back of the larynx. The Adam’s apple, the bump that is especially visible on the front of the throat of men, is the protrusion of the thyroid cartilage. The lower pair of vocal folds, whose inner edges are attached to the vocal ligaments, can be pulled apart or brought together by movement of the arytenoid cartilages. When the vocal folds are spread apart (as in figure 1.2) and the airstream passes freely through the space between them (the glottis), the sound that is produced is characterized as voiceless. The sound you produce when you pronounce the in the word ‘finger’, for example, is voiceless. When the vocal folds are brought together and the air passing through them causes them to vibrate, the resulting sound is characterized as voiced. All the sounds in the word ‘lion’ are voiced. You can determine whether a sound is voiced or voiceless by placing a finger or two on the front of your throat (on your Adam’s apple) and saying a sound. You should be able to feel your vocal folds vibrating when you produce the sounds represented by the following orthographic symbols: in ‘yes’; in ‘already’; in ‘good’; in ‘mother’. Your vocal folds should not vibrate when you produce in ‘it’; in ‘I’; or in ‘shoe’. After the airstream leaves the larynx, it travels into the vocal tract for the final phase of speech production, articulation. The vocal tract is the upper respiratory

Figure 1.2 The larynx. Modified from I. R. Titze, “Principles of Voice Production: Second Printing.” Copyright 2000 by the National Center for Voice and Speech, Iowa City, IA 52242. Courtesy of Ingo R. Titze.

6  Phonetics tract, the area above the glottis that includes the oral cavity (the mouth), the nasal cavity, and the pharynx (the part of the throat between the oral and nasal cavities and the larynx). If the airstream passes solely through the oral cavity after it passes through the glottis, the resulting sound is characterized as being oral. The velum, the soft area at the very back of the roof of the mouth, plays an important role in the production of oral sounds. When the velum is raised toward the pharynx (the back of the throat), this creates a barrier to the nasal passage and allows air to be released only through the mouth. Most of the sounds of German are oral. All of the sounds in the word ‘love’, for example, are oral. If you lower your velum, allowing air to pass through the nasal cavity, the sound you produce is characterized as being nasal. The word ‘mouth’ contains two of the three nasal sounds in German—the sounds represented orthographically by and . There are other organs of speech in addition to the velum that play an active role in the articulation of speech. If you look in a mirror, open your mouth, and say “ahhh,” you can see a small piece of soft tissue hanging down from the back of your velum. This is the uvula. The uvula plays a role in producing two of the sounds that are acceptable ways of pronouncing the in a word like ‘red’. The lips, which are very flexible, are also active articulators. They can be used together in the creation of a speech sound. The sound represented by in ‘oven’, for example, is made with the lips in a rounded position. The lips can also be used with the teeth to produce speech sounds. The sound of in ‘celebration’, for example, is produced with the lower lip and upper front teeth. The primary active articulator and the most flexible one is the tongue. It can be raised, lowered, pushed forward, pulled back, and so on, so that it approaches or touches other surfaces in the mouth. Slowly say the word ‘same’ and feel your tongue move and notice how different parts of it touch different surfaces in your mouth. The specific areas of the tongue that are identified when describing the role that the tongue plays in producing speech sounds are shown in figure 1.3. Because the tongue does not have any obvious natural divisions, these areas are typically described in terms of their position relative to other parts of the mouth when the tongue is at rest. The back of the tongue lies below the velum. The front of the tongue lies below the hard palate, the roof of the mouth. The blade of the tongue lies below the alveolar ridge, the area right behind the upper teeth. The tip of the tongue is the very end of the tongue. To describe the role that the tongue plays in producing the sound for the in the word ‘pine’, for example, we can say that the blade (or the tip) of the tongue is pressed against the alveolar ridge. Say the word and notice what you do with your tongue when you make this sound. Did you notice that you also put the blade (or tip) of your tongue against the alveolar ridge when you pronounce the sound for ? If you say the word ‘can’, you will notice that the back of the tongue makes

Phonetics  7

t fron

back

d bla

e

tip

Figure 1.3 The tongue (adapted from International Phonetic Association 1999:7).

contact with the velum when you make the sound for . When we discuss the sounds of German in more detail, you will see that the position of the tongue plays an important role in characterizing and distinguishing the different sounds. 1.5. Classifying Sounds Speech sounds can be grouped into sound classes on the basis of the phonetic characteristics they share. We have already talked about two different groupings. All voiced sounds, for example, form a class—in contrast to voiceless sounds, which also form a class. Another grouping is the class of oral sounds, which contrasts with the class of nasal sounds. The most basic division of sounds is the distinction between consonants and vowels. This distinction can be expressed in terms of the extent to which the airstream is obstructed or impeded during the production of the sound. Consonants are those sounds that are produced by obstructing the airstream in some way as it flows through the vocal tract. The sounds represented in German by

, , , , and , for example, are all consonants. Vowels, on the other hand, are produced with little or no obstruction of the airstream as it flows through the vocal tract. The different vowel sounds are produced by changing the position of the tongue and the lips. Both consonants and vowels can be further classified. There are different ways in which the airstream can be obstructed, and therefore different classes of consonants that can be identified on the basis of how the airstream is impeded. Vowels can also be grouped into classes depending on parameters such as the position of the tongue and whether the lips are rounded or unrounded. We will discuss the classification of consonant sounds first, since it is easier to feel the movements of the tongue and contacts of the articulators when these sounds are produced.

8  Phonetics 1.5.1. Classifying Consonants For the purpose of classifying the consonants of German, we will need to consider four factors, all of which involve different ways in which the airstream is modified after it exits the lungs: 1) the state of the vocal cords, that is, whether they are together and vibrating, in which case the consonant is described as being voiced, or apart and not vibrating, in which case the consonant is characterized as voiceless; 2) the position of the velum, either raised and therefore sending the airstream solely through the mouth to produce an oral sound, or lowered and allowing air to escape through the nose, producing a nasal sound; 3) the place of articulation, that is, the location of the obstruction of the airstream; and 4) the manner of articulation, or the way in which the airstream is obstructed. Since we have already discussed the first two factors (the role that the vocal cords and the velum play in the production of a consonant) we will concentrate here on the last two—place of articulation and manner of articulation. There are eight places of articulation that are necessary for the description and classification of the consonant sounds in German. If we begin at the front of the mouth, the first important articulators are the lips. A consonant that is produced by obstructing the airstream with both lips is characterized as bilabial (for example, the sound for in the word ‘bone’). If a sound is produced using the lower lip and upper incisors to impede the airstream, the sound is labiodental (for example, the sound in ‘fist’). If the tip or blade of the tongue touches or comes close to the alveolar ridge in order to restrict the airflow, the resulting sound is alveolar (such as the and sounds in ‘leather’ and the sound in ‘house’). If the tongue touches or comes close to the area slightly behind the alveolar ridge, the resulting sound is postalveolar (such as the sound represented by in ‘shoe’). If the front of the tongue is on or near the hard palate during the production of a consonant, it is characterized as palatal (for example, the sound in ‘young’ and the sound represented by in ‘break’, known as the “ich-Laut”). If the back of the tongue touches or is near the velum, the sound produced is velar (such as the sound in ‘can’, as well as the sound represented by in ‘book’, known as the “ach-Laut”). When a sound is produced by raising the back of the tongue toward the uvula, it is characterized as uvular. One example of a uvular sound is the “Zäpfchen-R” ‘uvular r’, which is articulated by causing the uvula to vibrate against the raised back of the tongue. The uvular is one possible pronunciation of the in a word like ‘red’. The final place of articulation relevant for German is the glottis, the space between the vocal cords. An articulation that involves the glottis is characterized as glottal. The in words like ‘help’ is glottal. The glottis is open when it is produced.

Phonetics  9 When we consider manner of articulation, that is, the way in which the airstream is obstructed when a consonant is produced, we need to recognize five different types of consonants in order to classify the consonant sounds of German: plosives, nasals, trills, fricatives, and approximants. Plosives are produced by a complete, momentary closure of the vocal tract and subsequent release. Because pressure builds up during closure, this release is accompanied by an “explosion” of sound (hence the term “plosive”). Plosives are also known as stops. The

sound in ‘passport’, for example, is a plosive, [p]. It is voiceless (the vocal folds are apart and do not vibrate during closure) and bilabial (the closure is formed with both lips). In addition to voiceless plosives, German also has voiced plosives, which are produced with the vocal folds together and vibrating during closure. The sound in ‘above’ is a voiced bilabial plosive, [b]. Nasals are similar to plosives in that a complete closure is formed in the vocal tract. They differ from plosives, however, in that the velum is lowered and the air is allowed to escape through the nasal passage. The sound in ‘mother’, for example, is a bilabial nasal, [m]; the closure is formed with both lips. All nasals in German are voiced. Trills are produced by placing one articulator close to another articulator and holding it there loosely so that the airstream causes it to vibrate. There are two trills in German. The alveolar trill, [r] (das Zungenspitzen-R), is an sound typically found in Bavaria and Austria. It is produced with the tip of the tongue vibrating against the alveolar ridge. The uvular trill, [ʀ] (das Zäpfchen-R), mentioned above, is the second kind of trill in German. Both the alveolar and the uvular trill are voiced. Fricatives are consonants that are produced when two articulators are brought close together, creating a narrow passage that causes turbulence in the air passing through. The sound in ‘fever’, for example, is a fricative. The lower lip and the upper incisors are brought together to create the narrow opening that produces a turbulent airflow. This sound is voiceless (the vocal cords do not vibrate when it is produced); it is a voiceless labiodental fricative, [f ]. German has voiced as well as voiceless fricatives. The voiced counterpart to [f ] is [v], a voiced labiodental fricative. An example of this sound is the in the word ‘to wash’. Approximants are produced by bringing two articulators close together, but not close enough to cause a turbulent airflow. The initial consonant in the word ‘year’ is a voiced palatal approximant, [ j]. German also has a lateral approximant, [l], which is produced by placing the blade of the tongue against the alveolar ridge and allowing the air to escape on one or both sides of the tongue. This sound is voiced and is classified as alveolar, since the point of contact is the alveolar ridge.

10  Phonetics Thus far we have talked about simple consonant sounds—those that can be represented using a single IPA symbol. There is another set of consonants that are more complex and cannot be represented with a single IPA symbol: affricates. We will define an affricate as a plosive followed by a homorganic fricative—a fricative with roughly the same place of articulation. The word ‘tin’, for example, begins with an affricate. The in this word represents a [t] (an alveolar plosive) followed by an [s] (an alveolar fricative): [tsɪn]. 1.5.2. Classifying Vowels Vowel articulations are more subtle than those for consonants and are not as easy to feel, since your tongue does not normally make contact with the roof of the mouth when you make a vowel sound. However, you will find that vowels are not as mysterious as you may have thought. There are five parameters that we will use to classify and describe the vowel sounds of German, and once you under­ stand how these work, much of the mystery of these sounds should disappear. These five parameters are tongue height, tongue position, lip rounding, tenseness, and length. All vowels in German are voiced (vowels are generally voiced), so we do not need a parameter of voicing in our classification of the different vowel sounds of German. The IPA system of vowel classification is somewhat different from the one we will use here, but it is useful for providing a visual representation of where the different vowel sounds in German are produced in the mouth. The chart in figure 1.4 contains the IPA symbols for the vowel sounds that occur in German. The quadrilateral figure represents the space in the mouth within which the tongue can move to produce vowel sounds. Just imagine a side view of an open mouth

Front i y

Central

I

Close-mid

e

Back u Ω

Close

Y

o

ø e ε

œ

c

Open-mid

a Open

aː a

Figure 1.4 The IPA quadrilateral for vowels in German (adapted from Hall 2000:34).

Phonetics  11 (with the mouth opening on the left) superimposed over this vowel quadrilateral. The place in the mouth where the tongue is as far forward and as high as possible is where the [i] sound is made, the sound for in a word like ‘style’. The [u] sound is made with the tongue raised as high as possible at the back of the mouth (and the lips rounded). This is the sound for in a word like ‘you’. When we present all the vowel sounds of German in the following section, we will use a chart that is similar to the chart in figure 1.4. The [i] sound, for example, will be in the upper left-hand corner of the chart, and the [u] sound will be in the upper right-hand corner. It will be more abstract than the chart in figure 1.4, however, since it will not represent the actual position of the vowel sounds. In addition, the IPA system of vowel classification uses four tongue heights to distinguish the vowels of the world’s languages, as you see in figure 1.4 (close, close-mid, openmid, and open), whereas we will use a slightly different system of three tongue heights (high, mid, and low) together with the parameter of tenseness to characterize the vowels of German. There is a correlation between the tongue height of a vowel and the degree of mouth opening when that vowel is articulated. The sound in ‘look’, [iː], is a high vowel; the mouth is almost closed when this vowel is pronounced. (The symbol [ː] indicates that this vowel is also long. More will be said about vowel length below.) The sound in ‘lake’, [eː], is a mid vowel; the mouth is opened slightly (the jaw is lowered somewhat) to produce this vowel. The sound in ‘saw’, [aː], is a low vowel; the mouth is opened wide to produce this sound (the jaw is very low). If you say these vowels in succession—[iː], [eː], [aː]—you can feel your jaw lowering. As noted above, there are essentially three degrees of tongue height necessary to distinguish the different vowel sounds of German: high, mid, and low. If you keep in mind the correlation between tongue height and how wide your mouth is open when you pronounce a vowel, this will help you have a better feel for the parameter of tongue height. The parameter of tongue position identifies the position, from front to back, of the highest part of the tongue. The sound in ‘song’, [iː], is a front vowel. The highest part of the tongue is in the front of the mouth when [iː] is articulated. In contrast, the sound in ‘good’, [uː], is a back vowel. The highest part of the tongue is in the back of the mouth when [uː] is pronounced. If you say [iː] and then [uː], you can feel your tongue retracting for the pronunciation of [uː]. Your lips also become rounded when you say [uː], but ignore this and concentrate on what your tongue is doing; you will feel your tongue pulling back when you go from [iː] to [uː]. The sound in ‘yes’, [aː], is a central vowel; the highest part of the tongue is roughly in the center of the mouth along the front-to-back dimension. If you say [iː] and then [aː], you will feel your tongue move back a bit in your mouth. Your mouth will also open wider (your jaw and your tongue will lower) when you go from [iː] to [aː], but focus on what your tongue is doing in

12  Phonetics terms of its movement from front to back in your mouth. There are only three tongue positions that are necessary to characterize German vowels: front, central, and back. The parameter of lip rounding involves the position of the lips. If the lips are rounded in the articulation of a vowel, the vowel is characterized as rounded. If not, the vowel is unrounded. Lip rounding is probably the easiest parameter to work with. You can simply look in a mirror and see what your lips are doing when you make vowel sounds. If you say the sound in ‘you’, [uː], you will see that your lips are rounded. If you say the sound in ‘the’, [iː], your lips will be spread apart. An interesting fact about German as compared to English is that it has front vowels that are rounded, which English does not. The sound [ yː] in the word ‘sweet’, for example, is a high front rounded vowel. This sound does not exist in English. However, we can easily demystify it by pointing out that the only difference between [ yː] and [iː] is one of rounding. If you say and hold the [iː] sound and then round your lips without moving your tongue, you will be pronouncing the sound [ yː]. Tenseness is commonly described as the amount of muscular tension necessary for the production of a vowel. A tense vowel is characterized as being produced with more muscular tension than a lax vowel. Although no experiments have demonstrated a difference in muscular tension, there are phonetic differences between tense and lax vowels. There are a number of pairs of tense and lax vowels in German, [iː]/[ɪ], [eː]/[ɛ], and so on. The tense members of these pairs are located along the outer edges of the vowel space, whereas the lax members are closer to the center. For example, there are two sounds in German: the tense (and long) [iː] in ‘crooked’ and the lax (and short) [ɪ] in ‘ship’. If you look at the position of these two sounds in the vowel quadrangle in figure 1.4, you will see that [iː] is located higher and further forward in the mouth than [ɪ]. The final parameter necessary for the description of vowel sounds in German is length. Length can be characterized as the relative duration of a vowel in comparison to other vowels. The sound in ‘state’, [aː], is long; the in ‘city’, [a], is short. As you saw above, the sound in is long (as well as tense), [ʃiːf ]; the in is short (as well as lax), [ʃɪf ]. The long [aː] and short [a] do not differ in tenseness. They only differ in length. German orthography provides important information about vowel length. This will be discussed in chapter 8. The vowels we have talked about up to now are monophthongs—simple vowels produced with no change in quality (tongue height, tongue position, lip rounding) within a syllable. A diphthong, in contrast, is a vowel that does change in quality during a syllable; the tongue moves during the articulation of the vowel and the lips may change position as well. If you say the word ‘am’, for example, and

Phonetics  13 then the word ‘leg’, you will notice that your tongue does not move when you pronounce the vowel in , [ɪ], but it does when you pronounce the in . A common way of representing diphthongs phonetically is by using two vowel symbols. The first vowel represents the starting point of the articulation of the diphthong; the second vowel represents the endpoint. One of the sounds in a diphthong is more prominent than the other (it is longer, for example), and this is indicated by putting the diacritic [ ̯ ] under the less prominent sound. The diphthong in the word is represented as [aɪ ̯]. When this diphthong is produced, the tongue starts in the low, central position for [a] and moves forwards and upwards to the high, front position for [ɪ]; the diacritic [ ̯ ] is placed under [ɪ], which is the less prominent vowel sound in the syllable. When the diphthong in ‘house’, [aʊ̯], is produced, the tongue moves from the low, central position for [a] to the high, back position for [ʊ] and the lips become rounded. 1.6 . The Phones of German Now that you have been introduced to the ways in which consonants and vowels are classified and described in general, we will look in more detail at the consonant and vowel sounds in German. The following two sections present the relevant articulatory characteristics of all the speech sounds, or phones, of German. All the sounds of German are presented in appendix A with examples of words that contain them. Recordings of these words can be found on the Web site for this book. 1.6 .1. The Consonants Table 1.1 presents all the consonant sounds of German, beginning with the plosives at the top (the consonants that involve the greatest obstruction of the airstream) and ending with the approximants (the consonants that involve the least amount of obstruction). The left-hand column in the table lists all the types of consonants that are found in German, that is, all the different manners of articulation that are relevant. (Lateral sounds are conventionally listed separately from their nonlateral counterparts, so the table includes a separate row for the lateral approximant, [l].) The row at the top of the table lists all the relevant places of articulation for the consonants in German. There are two types of plosives in German—voiceless and voiced—and there are three main places of articulation for each type: bilabial, alveolar, and velar. These places of articulation locate the closure in the vocal tract that is part of the production of the plosive. There is also a voiceless glottal plosive, [ʔ], most commonly called the glottal stop, which is produced by holding the vocal cords tightly together so that the closure is formed at the glottis. We make glottal stops in English when we say a word like ; the glottal stop is the sound that occurs before each vowel in this word.

Table 1.1 The consonant sounds of German. Voiceless sounds are on the left under each place of articulation; voiced sounds are on the right. Plosive Nasal Trill Fricative Approximant Lateral  approximant

Bilabial

p

b m

Labiodental

f

v

Alveolar

t

s

d n r z l

Postalveolar

ʃ

ʒ

Palatal

ç

k

j

x

Velar

g ŋ

Uvular

Glottal Ɂ

ʀ ʁ

h

Phonetics  15 (1)

Voiceless plosives [p] ‘breakdown’ [panə]; ‘lip’ [lɪpə]; ‘type’ [tyːp] [t] ‘part’ [taɪ ̯l]; ‘please’ [bɪtə]; ‘with’ [mɪt] [k] ‘account’ [kɔnto]; ‘jacket’ [ jakə]; ‘happiness’ [ glʏk] [ʔ] ‘out’ [ʔaʊ̯s]; ‘to end’ [bəʔɛndən]

Voiced plosives are more restricted than voiceless plosives. Unlike voiceless plosives, they can never occur at the end of words or syllables. (2)

Voiced plosives [b] ‘ball’ [bal]; ‘but’ [ʔaːbɐ] [d] ‘roof ’ [dax]; ‘or’ [ʔoːdɐ] [ g] ‘god’ [ gɔt]; ‘camp’ [laːgɐ]

If you see , , or at the end of a word, for example, you typically pronounce them like you would pronounce their voiceless counterparts,

, , and : ‘kind’ [liːp]; ‘mouth’ [mʊnt]; ‘likes’ [maːk]. This process is referred to as final devoicing, and it will be taken up in later chapters. The nasals in German, which are always voiced, have the same places of articulation as the three main plosives: bilabial, alveolar, and velar. (3) Nasals [m] ‘mouth’ [mʊnt]; ‘heaven’ [hɪməl]; ‘tree’ [baʊ̯m] [n] ‘nose’ [naːzə]; ‘under’ [ʔʊntɐ]; ‘Vienna’ [viːn] [ŋ] ‘to sing’ [zɪŋən]; ‘tight’ [ʔɛŋ]

There are two trills in German, an alveolar trill, [r], and a uvular trill, [ʀ], both of which are voiced. These trills are two of the ways in which can be pronounced in German. (4) Trills [r] ‘role’ [rɔlə]; ‘lesson’ [leːrə]; ‘green’ [ gryːn] [ʀ] ‘role’ [ʀɔlə]; ‘lesson’ [leːʀə]; ‘green’ [ gʀyːn]

The choice among these and other pronunciations of is dependent on a number of factors—its location in a word, the regional variety of German that a speaker uses, and so on. We will talk more about the pronunciation of in later chapters. There are more places of articulation for fricatives in German than for stops and nasals: labiodental, alveolar, postalveolar, palatal, velar, uvular, and glottal. There are both voiced and voiceless fricatives, although not necessarily at each place of articulation.

16  Phonetics (5)

Voiceless fricatives [f ] ‘fat’ [fɛt]; ‘oven’ [ʔoːfən]; ‘run’ [laʊ̯f ] [s] ‘scale’ [skaːla]; ‘water’ [vasɐ]; ‘house’ [haʊ̯s] [ ʃ ] ‘beautiful’ [ʃøːn]; ‘bottle’ [flaʃə]; ‘meat’ [flaɪ ̯ʃ] [ç] ‘China’ [çiːna]; ‘surface’ [flɛçə]; ‘I’ [ʔɪç] [x] ‘to make’ [maxən]; ‘book’ [buːx] [h] ‘today’ [hɔɪ ̯tə]; ‘secret’ [ gəhaɪ ̯m]

Voiced fricatives, like voiced plosives, never appear in word-final or syllable-final position. (6)

Voiced fricatives [v] ‘when’ [van]; ‘lion’ [løːvə] [z] ‘lake’ [zeː]; ‘to sneeze’ [niːzən] [ʒ] ‘genius’ [ʒeniː]; ‘floor’ [ʔetaːʒə] [ʁ] ‘role’ [ʁɔlə]; ‘lesson’ [leːʁə]

Notice that the voiced uvular fricative, [ʁ], is another possible pronunciation of the sound in German. There are two approximants in German, one of which is [ j], a voiced palatal approximant. (7) Approximants [ j] ‘yes’ [ jaː]; ‘majesty’ [majɛsteːt]

The second approximant is the lateral approximant, [l], which is alveolar. (8)

Lateral approximants [l] ‘to laugh’ [laxən]; ‘school’ [ʃuːlə]; ‘full’ [fɔl]

The lateral approximant is also voiced. Not included in table 1.1 are the four affricates that can be recognized for German. The chart is not set up to represent sequences of consonants with different manners of articulation, so affricates do not lend themselves to this type of display. (9) Affricates [pf ] ‘pound’ [pfʊnt] [ts] ‘to’ [tsuː] [tʃ] ‘bye’ [tʃʏs] [dʒ] ‘jungle’ [dʒʊŋəl]

Notice that [ts] is the only affricate in which the plosive and fricative have the exact same place of articulation. In [pf ], for example, [p] is bilabial, whereas [f ] is labiodental.

Phonetics  17 1.6 .2. The Vowels Table 1.2 presents all the vowel sounds of German using the parameters of tongue height (high, mid, low), tongue position (front, central, back), lip rounding (unrounded, rounded), tenseness (tense, lax), and length. Table 1.2 differs from a typical IPA chart for vowels (see the chart in figure 1.4, for example) in that the placement of the vowels does not depict the actual (or approximate) position of the tongue when these vowels are pronounced. However, the table can be viewed as representing the position of these vowels in very general terms. The vowels listed as front vowels are made at the front of the oral cavity, the vowels listed as back vowels are made at the back, and so on. (We are assuming the standard representation of the oral cavity, with the mouth opening to the left.) The table provides a convenient way of supplying the relevant articulatory characteristics of each vowel sound in German. For example, the table tells us that [œ] is a mid, front, lax, short, rounded vowel. The front vowels are particularly interesting because for each unrounded vowel there is a rounded counterpart: (10) Front vowels [iː] ‘bee’ [biːnə] [ɪ] ‘box’ [kɪstə] [eː] ‘to read’ [leːzən] [ɛ] ‘to know’ [kɛnən]

[ yː] [ʏ] [øː] [œ]

‘stage’ [byːnə] ‘coast’ [kʏstə] ‘to solve’ [løːzən] ‘to be able to’ [kœnən]

Note that each unrounded front vowel in German differs from its rounded counterpart only in the feature of lip position. For example, [ɛ] is a mid, front, lax, short, unrounded vowel; its rounded counterpart, [œ], is a mid, front, lax, short, rounded vowel. There are four central vowels in German, two of which never occur in stressed syllables: [ə] (schwa) and [ɐ] (dark schwa). These two are mid, central, lax, short, unrounded vowels. [ɐ] is somewhat lower than [ə] and could more accurately be described as a lower-mid, central vowel. Table 1.2 The vowel sounds of German. Front

High Mid Low

Tense Lax Tense Lax

Unrounded iː ɪ eː ɛ

Rounded yː ʏ øː œ

Central Unrounded

ə ɐ aː a

Rounded

Back Unrounded

Rounded uː ʊ oː ɔ

18  Phonetics

Substitute [eː] for [ɛː] There is one vowel in German that we have not included in table 1.2, [ɛː]. It is a long version of the vowel in a word like ‘bed’ [bɛt]. The [ɛː] sound is typically given as the pronunciation of in words like ‘bear’, ‘cheese’, and ‘to count’. This is the sound you will find if you look up these words in the Duden pronouncing dictionary, Duden: Aussprachewörterbuch (Mangold 2005). We recommend instead using [eː], which is an alternative pronunciation that is fully acceptable in the standard language (Mangold 2005:21). This is the pronunciation we use in our transcriptions. For further information on [ɛː], see the discussion of vowels in the advanced topics section of this chapter.

(11)

Mid, central vowels [ə] ‘role’ [ʁɔlə] [ɐ] ‘scooter’ [ʁɔlɐ]

Schwa is the pronunciation of an unstressed at the end of a word. In adjective and verb endings, for example, is pronounced as schwa ( ‘cold milk’ [kaltə mɪlç]; ‘I’m coming’ [ʔɪç kɔmə]). An example of [ɐ] is the pronunciation of the sequence at the end of a word ( ‘driver’ [faːʁɐ]; ‘children’ [kɪndɐ]; ‘better’ [bɛsɐ]). The remaining central vowels are both low, [a] and [aː], and differ from each other only in terms of length.

Pronunciation of the ending You may have noticed that native speakers of German often pronounce the ending by omitting the vowel, schwa, and using the nasal, [n], to “fill out” the syllable. Some examples are pronunciations like [vɪsn̩] for ‘to know’ and [bluːtn̩] for ‘to bleed’ (the diacritic [ ]̩ indicates that [n] is syllabic or ­syllable-forming). We have similar pronunciations in English, for example, [lɪsn̩] for and [kɑtn̩] for . This “deletion” of schwa in German is common in the ending after a fricative, but it can also be found before other consonants (before [m] and [l], for example). Pronunciations with schwa deletion are normal unless a speaker is speaking slowly and clearly. In chapters 6 and 13 we discuss in more detail the conditions under which schwa is deleted, and in chapter 13 we provide you with the opportunity to practice schwa deletion in your own speech. In general (with the exception of our discussion of schwa deletion in chapters 6 and 13), we include schwa in our transcriptions of words with the ending (and those with and as well), and represent words like and phonetically as [vɪsən] and [bluːtən].

Phonetics  19 (12) Low, central vowels [a] ‘city’ [ʃtat] [aː] ‘state’ [ʃtaːt]

[a] is a low, central, short, unrounded vowel; [aː] is a low, central, long, unrounded vowel. Neither is marked for tenseness; we treat them here as being neither tense nor lax. Examples like those in (12) show us the importance of distinguishing German vowels for length. There are four back vowels in German, all of which are rounded. (13) Back vowels [uː] ‘to haunt’ [ʃpuːkən] [ʊ] ‘to spit’ [ʃpʊkən] [oː] ‘oven’ [ʔoːfən] [ɔ] ‘open’ [ʔɔfən]

Notice that each of these sounds has a front rounded counterpart. (14) Rounded vowels [uː] ‘brother’ [bʁuːdɐ] [ʊ] ‘mother’ [mʊtɐ] [oː] ‘oven’ [ʔoːfən] [ɔ] ‘to require’ [fɔɐ̯dɐn]

[ yː] [ʏ] [øː] [œ]

‘brothers’ [bʁyːdɐ] ‘mothers’ [mʏtɐ] ‘ovens’ [ʔøːfən] ‘to support’ [fœɐ̯dɐn]

The only difference between [ʊ] and [ʏ], for example, is that [ʊ] is back and [ʏ] is front. Both are otherwise high, lax, short, rounded vowels. If you say [ʊ] and then [ʏ], you can feel your tongue move forward for the pronunciation of [ʏ]. In addition to the vowels in table 1.2, German has the diphthongs [aɪ ̯], [aʊ̯], and [ɔɪ ̯]. They are not included in the chart because it is suitable only for characterizing simple vowel sounds. Remember that diphthongs are complex vowel sounds that change in quality during the syllable in which they are articulated. (15) Diphthongs [aɪ ̯] ‘fine’ [faɪ ̯n] [aʊ̯] ‘mouse’ [maʊ̯s] [ɔɪ ̯] ‘joy’ [fʁɔɪ ̯də]

Notice that when is “umlauted” to to make the plural form of a noun, for example, this results in a change from the diphthong [aʊ̯] to the diphthong [ɔɪ ̯]: ‘house’ [haʊ̯s], ‘houses’ [hɔɪ ̯zɐ]. As the word shows, the sequence is also used to represent the diphthong [ɔɪ ̯] 1.7 . Using IPA Symbols Two practical issues you will encounter when you use the IPA symbols in the charts in tables 1.1 and 1.2 are how to refer to these symbols and how to write

20  Phonetics them. Many of these symbols are identical to common orthographic symbols (regular letters), so you can refer to them just as you would refer to their orthographic counterparts. For example, you can refer to the symbol [n] as “lowercase N.” Other symbols have no orthographic counterparts in English or German, so these symbols will need special names. The symbol [ɔ], for example, is a special phonetic symbol, which we will call “open O.” The Handbook of the International Phonetic Association (International Phonetic Association 1999) lists the IPA symbols together with names that are for the most part those used in Pullum and Ladusaw’s Phonetic Symbol Guide (Pullum and Ladusaw 1986). We will use these names for the IPA symbols that appear in this book, and we have listed them in appendix B together with a brief phonetic description based on our method of describing the sounds of German. When you write IPA symbols, you must take care to write them just as they appear in print. One important point to pay attention to is the height of these symbols. In particular, the consonant and vowel symbols that are capital letters are written no higher than their noncapital counterparts. Small capital R and inverted small capital R, for example, both fill the same amount of vertical space as lowercase R: [ʀ], [ʁ], [r]. Similarly, small capital I fills the same amount of vertical space as lowercase I: [ɪ], [i]. The height of esh, [ ʃ ], the symbol for the sound often written as , is also important to note. This symbol is as high as the symbols [f ] and [l], but it also extends below the line of writing like the symbol [p]. You can see this in the transcription for the word ‘to splash’: [platʃən]. The symbols for small capital I, small capital Y, and upsilon ([ɪ], [ʏ], and [ʊ]) must be clearly distinguished from lowercase I, Y, and U ([i], [ y], and [u]). This is easy to do by making sure that you use serifs when you write [ɪ], [ʏ], and [ʊ]. Serifs are the short horizontal lines at the bottom and/or top of these letters. 1.8. Advanced Topics In this section we will provide additional information about the sounds of German in order to give you a more complete picture of some of the issues involved in describing and characterizing these sounds. 1.8.1. Consonants The glottal stop, [ʔ]—the sound that can be heard before the initial vowel in a word like —is typically classified as a glottal plosive and thus classified with [p], [t], and [k], which are plosives that are articulated in the vocal tract. However, from a purely phonetic standpoint, the glottal stop is a state of the glottis, one that can be described as a state in which the vocal cords are held together (Ladefoged 1971:16). Two other states of the glottis that we have talked about are voice (vibration of the vocal cords) and voiceless (vocal cords apart). We follow

Phonetics  21 customary practice and classify [ʔ] here as a glottal plosive. More information about the role this sound plays in German will be provided in chapter 3. Strictly speaking, [h] should not be classified as a fricative, because the noise associated with [h] is not the result of air being forced through a narrow gap. Instead, it is the result of air moving across the edges of the open vocal folds and other surfaces in the vocal tract; the sound of [h] can be likened to that of a noisy vowel (Ladefoged and Ferrari Disner 2012:57). Although it has been described as the voiceless counterpart of the vowel that follows it, the shape of the vocal tract during [h] is often that of the surrounding sounds (Ladefoged and Maddieson 1996:325). We follow common practice and treat [h] as a fricative. The sound found in words like ‘yes’, ‘now’, and ‘young’ is often characterized as a voiced palatal fricative (the IPA symbol for this sound is [ʝ]) rather than as a voiced palatal approximant, [ j]. It is pronounced with varying degrees of friction, which has no doubt led to these two different characterizations. According to Kohler (1995:156), when it appears after voiceless plosives and fricatives—in words like ‘well’ and ‘fiord’—it is at least partially devoiced and “fricativized.” Because it is generally realized as a voiced approximant, we treat it here as such. We characterize the “ach-Laut” as a voiceless velar fricative, [x], but it is also realized as a voiceless uvular fricative, [χ], which is produced by raising the back of the tongue toward the uvula rather than the velum (Kohler 1995:160–161). The uvular version of the ach-Laut is used after the vowels [aː], [a], and [ɔ]: ‘brought’ [bʁaːχ], ‘to make’ [maχən], ‘to cook’ [kɔχən]. The velar version is used after [uː], [ʊ], [o], and the diphthong [aʊ̯]: ‘cloth’ [tuːx], ‘ravine’ [ʃlʊxt], ‘high’ [hoːx], ‘also’ [ʔaʊ̯x]. To simplify the treatment of the ach-Laut, we use the symbol for the voiceless velar fricative, [x], to represent both versions. 1.8.2. Vowels Although [ɛː] is the pronunciation of that you will find in a dictionary for a word like ‘cheese’, it is not a pronunciation that is used in all varie­ ties of the standard language. Speakers in northern and eastern central Germany use the vowel [eː] instead. The use of [ɛː] as a pronunciation of is argued to have come about as a spelling pronunciation (Polenz 2000:151). Because words like ‘honor’ and ‘ear (of grain)’ were spelled differently (a result of having different histories), speakers felt they should be pronounced differently. (Words with used to be pronounced with an sound, and by using to spell them, their relationship to words that were still spelled with could be made clear. Compare ‘deed’ and ‘perpetrator’.) As you have seen, we use [eː] instead of [ɛː] in this book. Not only is it a fully acceptable way of

22  Phonetics ­ ronouncing in the standard language, it also allows us to simplify our prep sentation of the sounds of German. We use the parameter of tenseness to distinguish vowels like the [iː] in ‘crooked’ [ ʃiːf ] from [ɪ] in ‘ship’ [ʃɪf ]—[iː] is tense, whereas [ɪ] is lax. Other studies of German (Wiese 2000, for example) use the feature [ATR], which is the abbreviation for “advanced tongue root.” Vowels that have the feature [+ATR] involve a forward movement of the root of the tongue when the vowel is pronounced; these are vowels that we characterize as tense. Vowels that are [−ATR] do not involve this movement; these are vowels that we characterize as lax. Ladefoged and Maddieson (1990) argue, however, that the [ATR] feature that characterizes vowels in some African languages cannot be equated with tenseness in German vowels. All the vowels in table 1.2 are oral. German also has some nasalized vowels, which occur in loanwords that are typically from French. Nasalized vowels are produced by lowering the velum so that air escapes through the nose. The tilde, [˜], is used in phonetic transcription to indicate nasalization. (16) Nasalized vowels [ãː] ‘restaurant’ [ʁɛstoʁãː] [œ̃ ː] ‘perfume’ [paɐ̯fœ̃ ː] [õː] ‘balcony’ [balkõː]

Many words with nasalized vowels have alternative pronunciations in which the nasalized vowel is replaced by an oral vowel and a nasal consonant. (17) [ yːm] [ɔŋ], [oːn]

‘perfume’ [paɐ̯fyːm] ‘balcony’ [balkɔŋ], [balkoːn]

Nasalized vowels do not play a major role in German, so we will not treat them further in this book (beyond a brief discussion in the advanced topics section in chapter 3). Exercises 1.1. Think of the longest German word you know, and try to spell it. Try the same for the longest English word you know. How does the ease with which you spelled the words relate to the relative transparency of the orthographic systems of both languages? 1.2. As a word like ‘human life’ demonstrates, the relationship between the letters of the German alphabet and the sounds they represent is not always one-to-one. To become more aware of the differences between German spelling and pronunciation, do the following:

Phonetics  23 a. Find four words that show four different ways to spell the sound [f ]. b. Find five different examples of groups of two or more letters that represent only one sound. c. Explain how the in the word ‘tongue’ and the in the word ‘witch’ provide two examples of another way in which the correspondence between the letters of the German alphabet and the sounds they represent is not always one-to-one. 1.3. In a word like ‘please’, there are five different letters that represent four different segments (sounds): [bɪtə]. How many segments are there in each of the following words? a. b. c. d. e. f. g. h. i. j. 1.4. Say each of the following words. Is the first sound in each word voiced or voiceless? a. b. c. d. e. f. g. h. i. j. 1.5. Say each of the following words. Is the last sound in each word voiced or voiceless? a. b. c. d.

24  Phonetics e. f. g. h. i. j. 1.6. For each of the following pairs of sounds, pronounce the sounds, determine whether they have the same or a different place of articulation, and then identify the place of articulation for each sound. a. [p] : [m] b. [d] : [l] c. [t] : [ g] d. [f ] : [s] e. [k] : [x] f. [b] : [ŋ] g. [s] : [t] h. [p] : [v] i. [k] : [ŋ] j. [n] : [ ʃ ] 1.7. For each of the following pairs of sounds, pronounce the sounds, determine whether they have the same or a different manner of articulation, and then identify the manner of articulation for each sound. a. [k] : [ g] b. [f ] : [s] c. [ g] : [ŋ] d. [r] : [l] e. [v] : [ ʃ ] f. [n] : [z] g. [s] : [ç] h. [ʀ] : [ʁ] i. [m] : [ŋ] j. [ʔ] : [x] 1.8. Provide the phonetic symbols for all the consonants in German that are: a. voiceless plosives b. voiced plosives c. nasals d. trills e. voiceless fricatives

Phonetics  25 f. voiced fricatives g. approximants h. bilabial i. alveolar j. velar k. postalveolar l. labiodental 1.9. Provide the phonetic symbol for the following consonants in German: a. voiceless velar plosive b. voiced velar plosive c. velar nasal d. voiced bilabial plosive e. voiceless velar fricative f. voiceless palatal fricative g. voiceless alveolar fricative h. voiced alveolar fricative i. alveolar trill j. uvular trill k. lateral approximant l. voiceless postalveolar fricative m. alveolar nasal n. voiced palatal approximant o. voiced uvular fricative p. voiceless alveolar plosive q. voiced labiodental fricative r. voiceless glottal fricative 1.10. Match each of the following sounds with a word in the list below that contains that sound. a. [ j] b. [p] c. [z] d. [ŋ] e. [x] f. [v] g. [ ʃ ] h. [ç] , , , , , , , , ,

26  Phonetics 1.11. Describe the following consonant sounds phonetically (see the descriptions in exercise 1.9 above): a. [d] b. [m] c. [k] d. [v] e. [ʁ] f. [ j] g. [r] h. [ ʃ ] i. [ʔ] j. [s] 1.12. Pronounce each of the following vowels and determine whether it is rounded or unrounded. a. [aː] () b. [uː] () c. [ɛ] () d. [œ] () e. [ yː] () f. [eː] () g. [ʊ] () h. [ɔ] () i. [ɪ] () j. [ʏ] () 1.13. Pronounce each of the following pairs of vowels, determine whether they have the same or a different vowel height, and then identify the height of each vowel. Remember that there is a correlation between vowel height and the degree to which your mouth is open. Pay attention to the position of your lower jaw when you say these pairs of vowels. a. [iː] : [aː] (, ) b. [uː] : [ yː] (, ) c. [uː] : [oː] (, ) d. [aː] : [eː] (, ) e. [ʊ] : [ɔ] (, ) f. [ɔ] : [œ] (, ) g. [ɪ] : [ɛ] (, ) h. [ɔ] : [a] (, ) i. [oː] : [eː] (, )

Phonetics  27 1.14. Provide the phonetic symbols for all the vowels in German that are: a. high, tense vowels b. low vowels c. back vowels d. front, unrounded vowels e. high, front vowels f. mid, front vowels g. mid, central vowels h. front, rounded vowels 1.15. Provide the phonetic symbol for the following vowels in German: a. mid, back, lax, short, rounded vowel b. high, front, tense, long, unrounded vowel c. mid, front, lax, short, unrounded vowel d. low, central, short, unrounded vowel e. high, front, tense, long, rounded vowel f. lower-mid, central, lax, short, unrounded vowel g. mid, front, tense, long, unrounded vowel h. high, front, lax, short, rounded vowel i. high, back, tense, long, rounded vowel j. mid, front, lax, short, rounded vowel 1.16. Match each of the following sounds with a word in the list below that contains that sound. a. [aː] b. [ɛ] c. [ə] d. [ʊ] e. [ɔɪ ̯] f. [ʏ] g. [ɔ] h. [øː] , , , , , , , , ,

1.17. Describe the following vowels phonetically (see the descriptions in exercise 1.15): a. [ yː] b. [ɪ] c. [ʊ] d. [œ]

28  Phonetics e. [ɛ] f. [ə] g. [aː] h. [oː] i. [e:] 1.18. Using terms like “fricative,” “voiced,” “palatal,” “central,” and so on, provide one phonetic characteristic that all the segments in each group share and only these segments. a. [f v] b. [m n ŋ] c. [ə ɐ a aː] d. [ʃ ʒ] e. [uː ʊ oː ɔ] f. [a aː] g. [h ʔ] h. [r ʀ] i. [iː ɪ yː ʏ uː ʊ] j. [ɪ ʏ ʊ ɛ œ ə ɐ ɔ]

Chapter 2

Acoustic Phonetics

2.1. Introduction Have you ever noticed how difficult it is to understand what is being said when you are carrying on a conversation in a noisy room or on an airplane? Have you ever tried to understand heavily accented speech in that same noisy room or airplane? As a child, did you ever attempt to yell messages to your friends under­ water? Acoustic phonetics, the field of phonetics that deals with the transmission of speech sounds, can help us explain why it is more difficult to understand what people are saying in certain situations than in others. Accented speech can be difficult to understand in a quiet setting, and when additional noise is added, the task becomes even more challenging. In all of the situations above, it is difficult to understand the speech of others because the propagating medium (the substance that sound waves travel through) is less than ideal. In fact, in the case of water, it is impossible to understand anything that is being said, since sound is distorted as it travels through water. The field of acoustic phonetics has evolved over time. It is said that it started with the invention of the phonograph by Thomas Edison in the late nineteenth century. For the first time, people were able to record and analyze speech. In the mid-twentieth century, telephone companies were especially interested in  the transmission of speech, and companies like the Bell Telephone Company in the  United States and the Ericsson Telephone Company in Sweden established their own laboratories, which enabled a number of advances in acoustic phonetics. Telephone companies’ interest in acoustic phonetics was quite practical: researchers needed to make sure that listeners could understand what was being said on both ends of the telephone line. Acoustic phonetics has advanced as a discipline because researchers from a variety of fields—including acoustics, electronics, circuitry, computer science, phonetics, and phonology—have collaborated to better understand how speech is transmitted. Recent developments in computer technology mean that we are easily able to analyze finer details of the acoustics of speech. Today students and researchers alike are able to carry out sophisticated analyses of speech on their own computers by using readily accessible, free software. All of the waveforms and spectrograms in this book have been created in one such program, 29

30  Acoustic Phonetics Praat (Boersma and Weenink 2013). We are able to examine aspects of individual segments—like the noise bursts in plosives and the turbulent air in fricatives—as well as information in longer stretches of speech—like word and sentence stress and the intonation (melody) of utterances. When we analyze speech segments acoustically, we focus on the duration, quality, pitch, and loudness of sounds. The vowels of two very similar German words, ‘dimensions’ and ‘bulk’, differ along the first dimension, duration. The vowel in the first word takes more time to produce than the vowel in the second. In German, this difference is distinctive, that is, it distinguishes meaning. If you intend to say ‘bulk’ but produce a long instead of a short one, you will end up saying a word with a completely different meaning, ‘dimensions’. The vowels in the words ‘sons’ and ‘(I) long for’ differ in quality. In particular, they differ in terms of lip position. The first vowel in is produced with lip rounding and the first vowel in without. As these two words show, lip rounding is distinctive in German. Pitch is the sensation of a sound as being “high” or “low.” Pitch is distinctive in tonal languages like Chinese, where the same string of sounds can mean up to four different things depending on the pitch used to produce it. The acoustic measure that corresponds to pitch is fundamental frequency, which is the rate of vibration of the vocal folds. In German, pitch is not distinctive. This means that a word in German can be said with different pitches without changing its meaning. For example, the word ‘mother’ means the same thing, regardless of whether a speaker produces it with a high pitch, a low pitch, a rising pitch, or a falling pitch. Although pitch variation is not important for distinguishing the meanings of individual words in German, it does play an important role at the level of an utterance. For example, the changes in pitch over the course of an utterance, which we call intonation, can distinguish between questions and statements. You will learn more about intonation in chapter 5. Acoustic phonetics is an area that may be of special interest to you as a learner of German, as it is possible to analyze your own speech to determine the details about where and how you produce segments. You can also compare the details of your speech with those of a native speaker. It is important that you do not compare your own productions directly to those of only one native speaker, however. Anatomical differences (especially those between men and women), dialectal differences based on where a person comes from, and idiolectal differences among speakers (the idiosyncrasies of individual speakers) mean that no two individuals produce utterances in exactly the same way. In addition, the speech of nonnative speakers will show even greater variability depending, among other things, on first languages or first language dialects. Although the analysis of second language speech may appear to be a daunting task, if you keep in mind that speech production is always variable and if you remember important information about

Acoustic Phonetics  31 the distinguishing characteristics of various speech sounds, you will be able to make good use of acoustic analyses. This chapter is not intended to be a comprehensive introduction to acoustic phonetics. It is a basic introduction to be used with the exercises at the end of the chapter and in the second half of the book to give you a general idea about how you can use acoustic phonetics to understand more about German pronunciation. 2.2. Waveforms and Spectrograms It is easiest to measure sound transmission in a quiet room. One way of thinking about the transmission of sounds is to compare a quiet room to still water on a pond. If you drop a single pebble into the water, it sends a circle of waves across the water. Similarly, when any sound is produced, it sets the air particles in the quiet room in motion. The sound travels in waves across the room until it reaches our ears. The air particles crowd together and move apart again, creating changes in air pressure as the sound moves through the room. We can measure a number of aspects of the movement of these particles. The most important of these have to do with how often they repeat in a given amount of time (frequency) and the degree of movement of the particles (amplitude). You will encounter other terms that relate to particle movement, such as intensity, power, energy, sound pressure level, and loudness. For the purposes of this book, though, we will only deal with frequency and amplitude. The production of speech differs from that of other sounds. Consider a pure tone like the artificial beep sounds that are played when you have your hearing tested. The sound stays the same, regardless of how long it lasts, and we can see evidence of this when we analyze the sound’s waveform, that is, the graphic representation of a sound’s frequency (the number of repetitions per unit time) and amplitude. Waveforms are also referred to as oscillograms, but we will use the term “waveform” throughout this book. In the pure tone’s waveform in figure 2.1, we see that the signal repeats regularly. We refer to sounds with regularly repeating waveforms as periodic signals. Each repetition of the waveform is known as a cycle. One cycle is marked in figure 2.1. The regularity with which a waveform like that in figure 2.1 repeats affects the frequency of the sound. A sound that repeats more often during a given period of time has a higher frequency, and one that repeats less often has a lower frequency. The basic frequency at which a sound vibrates is known as its fundamental frequency, abbreviated F0. Frequency, which is the acoustic correlate of pitch, is measured in Hertz (Hz). The fundamental frequency of human speech is the rate of vocal fold vibration. If you take what you know about F0 and our perception of pitch, then you can predict that the speech of those people whose vocal folds vibrate more quickly (like young children and women) should be perceived as higher in pitch than the speech of those with vocal folds that vibrate more slowly. Human speech is more

32  Acoustic Phonetics

Figure 2.1 Waveform of a periodic signal.

complex than pure tones, however. During the production of a vowel, for example, sound vibrates in several ways at once, producing a periodic waveform that is complex. The waveforms of consonants are more varied than those of vowels, and this has to do with a consonant’s manner of articulation. Fricatives, for example, have waveforms that are aperiodic; they do not have a repetitive pattern. To explain the essential characteristics of speech acoustics, phoneticians have formulated a model known as the source-filter model. Fundamental frequency is a characteristic of the source, which is the larynx. Characteristics of vowels, nasals, approximants, and laterals known as formants (described below), as well as the specific patterns of the turbulent noise associated with fricatives and plosives, are a result of the filter—that is, the vocal tract. A waveform of human speech like the one in the upper half of figure 2.2 provides information about time on the x-axis (the horizontal axis) and about the air pressure on the y-axis (the vertical axis). The degree of change in air pressure when a sound is produced is known as its amplitude. The greater the change in air pressure, the greater the amplitude. You will not be able to determine everything that a person is saying from a waveform like the one in the top half of figure 2.2, but if you refer to the IPA transcription of the utterance, which is aligned with the segments in the waveform (and spectrogram in the bottom half), you should be able to interpret a number of salient features when you look at it. The vertical lines in the waveform correspond to the pulses produced by the vibrations of the vocal folds. You will note in the waveform that there are times when the vocal folds are not vibrating. For example, in the production of [k], a voiceless plosive, there is a silence that corresponds to closure associated with the plosive. You can observe this in the wave-

Acoustic Phonetics  33

Figure 2.2 Waveform and spectrogram of human speech. ‘The cat swims’.

form if you look at the relatively flat line that appears before the [k]. The same is true of the voiceless plosive [t]. A waveform also shows the difference between voiced sounds (those with waveforms that repeat regularly) and voiceless sounds (those with irregular fluctuations). The transitions into and out of sounds are also visible. You can see this most clearly in the vowels and nasals in figure 2.2, where you can see a buildup into and transition out of a series of waves with a relatively stable amplitude at their centers. From a spectrogram like that in the bottom half of figure 2.2, we can see some of the same characteristics. A spectrogram is a graphic representation of sounds in which time is shown on the x-axis, and frequency and amplitude are presented on the y-axis. In a spectrogram, you can see evidence of how and where vowels and sonorant consonants (nasals, [l], and [ j]) are produced. If you look at the vowels [iː], [a], [ə], and [ɪ] as well as the nasal [m], you can see dark horizontal bars throughout their duration. These are called formants, and their frequencies vary according to how and where speech segments are articulated. Because their values depend on how the speech stream is manipulated as it moves through the vocal tract, we refer to formants as resonant frequencies of the vocal tract. The

34  Acoustic Phonetics darkness of formant bands is indicative of amplitude. You will learn more about formants in the discussion of vowels and sonorant consonants below. Although we talk about speech as being made up of individual speech segments, it is important to remember that human speech is really continuous. If you refer again to figure 2.2, you will notice that there is no break between each of the speech sounds. Therefore, it is often difficult—especially when you first begin—to determine the boundaries of segments. It is important to keep this in mind as you continue your study of phonetics and phonology and as you transcribe speech into discrete symbols representing individual sounds. 2.3. The Acoustics of German Vowels When we produce vowels, the air is manipulated—but not obstructed—in various ways through the vocal tract. Remember that formants—the resonant frequencies of the vocal tract—provide information about where and how vowels are produced. They are characteristics of the filter. On a spectrogram like the one in figure 2.3, the formants are the dark bands corresponding to peaks of energy that are marked as F1 (first formant), F2 (second formant), and F3 (third formant). Vowels are largely characterized by the first two formants, which are the result of vocal tract shape as the vowels are being produced. F1 provides information about tongue height, and low vowels ([a]) have higher F1 values than do high vowels ([i], [ y], or [u]). F2 values give us information about tongue position, that is, the front-back dimension: front vowels ([i], [ y]) have higher F2 values than do back vowels ([u]). F2 values also provide information about an additional characteristic of vowels: the degree of lip rounding. Round vowels like [ y] and [ø] have lower F2 values than those vowels produced without lip rounding. Remember that formant values are expressed in Hz. The vertical striations (thin lines) in figure 2.3 correspond to vocal fold vibration. When we determine the formant values of vowels like those in figure 2.3, we rely on the center frequencies of F1 and F2. That is, we look for the steady-state portions of the vowels (those areas on the spectrogram in which the formants of vowels are relatively straight) and measure their formant values. One additional parameter to take into consideration when analyzing German vowels is duration. The vowel pair in ‘state’ [ ʃtaːt] and ‘city’ [ ʃtat], shown in figure 2.4, differs only in terms of duration. You can compare the duration of the vowels in the words by contrasting the size of the boxes. The vowel in is more than twice as long as the one in . Often vowels that differ in duration, like the sounds in ‘oven’ [ʔoːfən] and ‘open’ [ʔɔfən], also differ in tenseness. Because tense vowels are located along the outer edges of the vowel space and lax vowels are closer to the center, a tense vowel will be distinguished acoustically from its lax counterpart by having different F1 and F2 values.

Acoustic Phonetics  35

Figure 2.3 Spectrogram of German vowels [a], [u], [ y], and [i] with formants marked.

Figure 2.4 Spectrogram of German ‘state’ and ‘city’.

2.4. The Acoustics of German Consonants As you remember from chapter 1, we can distinguish among consonants in terms of manner of articulation: plosives, fricatives, nasals, approximants, and laterals. One additional distinguishing feature is voicing. Some German consonants are voiced; others are voiceless. Although it is possible to analyze other aspects of speech, these analyses go beyond the scope of what we will cover in this book. We can divide consonants into two general classes—obstruents and sonorants—

36  Acoustic Phonetics based on the degree to which the airflow is blocked as it moves through the articulatory tract. 2.4.1. The Acoustics of German Obstruents Obstruents—that is, those sounds that involve an obstacle or constriction in the vocal tract—can be classified according to the amount of blockage. They include plosives, fricatives, and affricates. They stand in opposition to sonorants, which are sounds that are more like vowels. Nasals, trills, [ j], and [l] are the sonorant consonants of German. All German sonorants are voiced, and obstruents can be voiced or voiceless. Plosives involve a complete blockage of the airstream. They are characterized by a brief silence due to this blockage followed by a burst of noise. This can be seen in the spectrogram of the word ‘boat’ in figure 2.5. There is a burst at the beginning, in the voiced plosive [b], and at the end, in the voiceless plosive [t]. Unlike plosives, fricatives do not involve a complete stoppage of the airflow. Instead, they are produced by constrictions in the vocal tract that lead to the generation of turbulent noise. In the spectrograms in figure 2.6, you can see evidence of this turbulent noise, which is the dark, irregular disturbance that is marked for each consonant. Unlike vowels, where the energy is densely organized into formant bands, the energy from fricatives is random and thus more similar to white noise. Nonetheless, as you can see from the location of irregular disturbance in figure 2.6, the frequency of the noise differs from one fricative to the next. It also varies by speaker. German is a special case, since it can be produced in a variety of ways: as a fricative, [ʁ], or a trill, [ʀ] or [r], when it occurs before a vowel in the same

Figure 2.5 Spectro­ gram of German ‘boat’.

Acoustic Phonetics  37

Figure 2.6 Turbulent noise in German fricatives.

Figure 2.7 Spectrogram of German ‘extension’.

­syllable; or as dark schwa, [ɐ̯], when it follows a vowel in the same syllable. If you look at the second in in figure 2.7, you will find evidence that it has been produced as the fricative [ʁ]. Here you can see evidence of the turbulent noise that is characteristic of fricatives. 2.4.2. The Acoustics of German Sonorants The sonorants of German include the nasals, [m], [n], and [ŋ]; the trills, [r] and [ʀ]; and the approximants, [ j] and [l]. Sonorants share a number of features with vowels. For example, they are always voiced and they exhibit formant structure. Importantly, we can distinguish nasals and laterals based on their formant ­patterns,

38  Acoustic Phonetics which provide important information about how these segments are articulated. We will not deal with [ j] here, as it is very similar to the vowel [i]. Because we focus on the fricative pronunciation of , we also do not deal with the trills. As you remember from chapter 1, air flows through the nose when we produce nasals. The nose adds volume to the vocal tract, and the formant bandwidths of nasals are wider than those of vowels and approximants. The walls of the vocal tract absorb the sound, and this leads to formants that are not as loud as they are in vowels. If you compare the [ŋ] in figure 2.7 to the vowels [ɛ] and [ə] on either side of it, you will note that although formants are visible, its formants are wider and lighter than those of the surrounding vowels. In laterals, the air flows out around the sides of the tongue. Laterals are like nasals in that the airflow is diverted without producing turbulence. Like nasals, [l] also exhibits formant structure. As you know, an additional distinguishing feature of German consonants is voicing. We can distinguish voiced from voiceless consonants by looking at the bottom of the spectrogram for what we refer to as the voice bar. Vocal fold vibration in vowels and in voiced consonants like [l] and [ŋ] can be seen in the dark bar at the bottom of the spectrogram in figure 2.7. You should also notice that the only voiceless sound in the entire word, [f ], does not exhibit the voice bar. 2.5. Advanced Topics Two important issues related to the production of plosives will be discussed in this section: aspiration and voice onset time (VOT). Aspiration is a burst of air that follows the German voiceless plosives, [p t k], most often when they are in word-initial position. The vocal folds do not vibrate when these consonants are produced, and there is a delay between the release of the closure and the onset of the vowel. During this time, a puff of air is released. You can feel the difference between aspirated and unaspirated plosives by producing the pairs of words in (1). If you hold the palm of your hand in front of your mouth when you say these words, you should be able to feel a puff of air after you produce the voiceless plosive at the beginning of the first word in each pair, but not after you say the voiced plosive at the beginning of the second word. (1)

‘agony’ [pʰaɪ ̯n] ‘tank’ [tʰaŋk] ‘checkout’ [kʰasə]

‘leg’ [baɪ ̯n] ‘thanks’ [daŋk] ‘alley’ [ gasə]

We indicate aspiration in phonetic transcriptions like those in (1) through the use of a superscript h, [ʰ], following voiceless plosives. Generally, though, we only indicate finer details like aspiration when we make use of narrow (detailed) transcriptions. That is, we indicate specific aspects of pronunciation that do not

Acoustic Phonetics  39 ­distinguish meaning only when we want to provide a more detailed transcription of the speech of an individual. VOT, also a quality of plosive consonants, is the length of time between the release of the closure (the burst) and the onset of voicing of the following sound. When we produce voiceless plosives with a great deal of aspiration, there is a significant period of time between the release of the closure and the beginning of the vowel that follows, so these plosives have a relatively long VOT. In fully voiced plosives, on the other hand, the vocal folds vibrate the entire time. Because the vocal cords start vibrating before the closure is released, these plosives have a negative VOT—a VOT that is less than zero. In reality, many of what we refer to as voiced plosives in German, especially those following voiceless sounds, are only slightly voiced or perhaps not even voiced at all. Therefore, we expect to be able to measure relatively long VOTs for voiceless aspirated plosives, and we are often also able to measure shorter VOT values for voiced plosives. We speak of VOT values in terms of duration (most often in milliseconds). It is important to remember that VOT values—like the degrees of aspiration—exist on a continuum, and they differ from one speaker (and even one utterance) to the next. VOT is marked in the words in figure 2.8.

Figure 2.8 Voice onset time in German ‘part’.

40  Acoustic Phonetics Exercises 2.1. Although we know that pitch and loudness are not distinctive in the production of German words, per se, there are instances in which pitch and loudness make a difference in how a listener interprets what is being said. Describe how pitch variation (for example, falling versus rising pitch) might affect how a child interprets the German word ‘now’ when it is spoken by a caregiver. Similarly, how might loudness also play a role in a child’s interpretation of what is being said? 2.2. The average first and second formant (F1 and F2) values for a subset of German vowels as produced by three males are provided in table 2.1. Enter these values in the vowel space in the chart below it. The values for [i] have been added for you. Be sure to label each of the vowels. How does the chart here correspond to the vowel chart provided in figure 1.4 in chapter 1? Table 2.1 Mean formant values for German vowels (data from Delattre 1965).

2500

2000

275 275 375 375 275 375 750

F2

2250 1750 2050 1600  850  850 1250

German Vowel Formant Values F2 1000 1500

500

0

0 100 200 300 400 500 600 700 800

F1

[i] [ y] [e] [ø] [u] [o] [a]

F1

Acoustic Phonetics  41 2.3. Imagine that you have been asked to record and analyze your own German vowels. Although you have been very careful in your analyses, you find that the formant values of your vowels are quite different from those in the table. Why might this be? 2.4. Refer to the waveform of at the top of figure 2.2. Considering what you know about manner and place of articulation as well as voicing, what are some of the sources of vibration in the consonants [d], [k], [s] and [ ʃ ]? Why do the vowels look different from the consonants in the waveform? 2.5. How does what you observed in the waveform in figure 2.2 correspond to what you know about obstruents versus sonorants? 2.6. Label each segment in the word in the spectrogram below as one of the following: vowel, plosive, or fricative.

2.7. Provide the phonetic transcriptions for three German words: , , and . The images below are the spectrograms of these words. Given what you know about acoustic phonetics, identify the word in each image. Then try to differentiate the segments in the spectrogram. Draw a box around each segment, and label it with the appropriate phonetic symbol.

This page intentionally left blank

Chapter 3

Phonology

3.1. Introduction If you ask typical native speakers of English about the sounds in the words and , they will tell you that the sounds are the same. Are they? You will probably be surprised to learn that they are not the same—at least not phonetically. The in is actually produced with the back of the tongue touching the roof of the mouth at a point that is further forward than for the in . If you put your tongue in a position to say and then try to say , you will discover that you want to move your tongue forward to get it in the right place to say the in . If we were to produce a very precise phonetic transcription of the in , we would use the diacritic [  ̟ ] (a subscript plus sign) to indicate that the [k] is advanced (produced further forward in the mouth than a [k] without this diacritic): [k̟iːp], [kaɪ ̯t]. What about the sounds in the words and ? They are clearly different, but not just phonetically. The in is a voiceless dental fricative, [θ]; the in is a voiced dental fricative, [ð]. The important difference, for speakers of English, is that these two sounds are capable of distinguishing meaning. If you substitute [θ] for [ð] in the word [ðaɪ ̯], you get a completely different word, a sequence of sounds with a completely different meaning, [θaɪ ̯]. Sounds like [θ] and [ð] are distinctive sounds in English. They contrast with other sounds and are capable of distinguishing meaning. In this chapter we will focus on the distinctive sounds in German—the sounds that native speakers perceive as different. We will show you how to determine which sounds are distinctive—not just in German, but in any language. We will also discuss some of the language-specific issues involved in identifying the set of sounds in German that are distinct. 3.2 . Basic Terms We call the distinctive sounds of a language phonemes, and we write them between slashes, //, to distinguish them from the phones (the actual sounds) of a language. In English, both /θ/ and /ð/ are phonemes. The phoneme /θ/ is realized (pronounced) as the phone [θ]; the phoneme /ð/ is pronounced as [ð]. Phonemes are the focus of the branch of linguistics called phonology. Whereas phonetics 43

44  Phonology deals with phones, phonology deals with phonemes—those sounds that are capable of differentiating meaning in a particular language. In addition, phonology studies how sounds alternate, how they replace one another in different forms of a word. For example, the alternation between the voiced and voiceless plosives [d] and [t] found in German words like ‘wheels’ [ʁeːdɐ] and ‘wheel’ [ʁaːt] is of interest to phonology. Syllable structure, stress, and intonation are also topics of phonology. In this chapter we focus on the phonemes of German. Other topics in the phonology of German (alternations, syllable structure, and so on) will be treated in later chapters. We will say a bit about syllables in this chapter, however, because they play an important role in our discussion of phonemes. How do we determine the phonemes of a language? That is, how can we figure out which sounds are distinctive? We start by looking for pairs of words with distinct meanings that differ by only one sound in the same position in each word. We call these minimal pairs. The words [θaɪ ̯] and [ðaɪ ̯], for example, form a minimal pair and show that the sounds [θ] and [ð] are capable of distinguishing meaning in English. The existence of a minimal pair like and allows us to posit /θ/ and /ð/ as phonemes of English. Other examples of minimal pairs in English are [pæt] and [bæt], [sɪn] and [sɪŋ], and [fiːt] and [fɪt]. Pairs like [tɪn] and [tæŋ] are not minimal pairs, however, because they differ in two sounds, the vowel and the final consonant. If we attempt to find a minimal pair in English that contrasts the sounds [k̟] and [k], we will come up empty-handed. To native speakers of English, these two sounds are the same—just two instances of the same phoneme, two ways of realizing the phoneme /k/. Native speakers automatically produce one or the other variant depending on where the sound occurs. They will use the more advanced [k̟] before front vowels like the [iː] in and the “normal” [k] before nonfront vowels like the [a] in the diphthong [aɪ ̯] in . These two sounds never occur in the same phonetic environment. Because they never occur in the same environment, we say that they are in complementary distribution. They are simply two variants (allophones) of the phoneme /k/. (This is a simplified description of the allophones of /k/ in English. See Cruttenden and Gimson [1994:153] for a more detailed treatment.) What is it about /θ/ in English that makes it a distinctive sound in comparison to /ð/? Both sounds are fricatives and both are dental (produced with the tip and rims of the tongue making contact with the edge of the upper incisors). The characteristic that distinguishes the two sounds is voicing: The phoneme /θ/ is voiceless; it has the feature [−voice]. The phoneme /ð/, in contrast, is voiced; it has the feature [+voice]. Because the feature of voicing is capable of distinguishing two phonemes, we say that it is a distinctive feature. Features are useful for characterizing classes of sounds. For example, the feature [rounded] can be used to identify

Phonology  45 the class of rounded vowels. The feature [tense] can be used to distinguish the tense vowels from their lax counterparts. Features are also useful for describing alternations. The feature [voice], for example, is useful for describing the alternation between voiced and voiceless plosives in German mentioned above. The feature [advanced] that characterizes the sound in , [k̟], is not a distinctive feature in English; it is a predictable feature. We know exactly when a sound in English will have an articulation that is advanced and when it will not. Because this feature is nondistinctive, we do not need to include it in our phonemic representation of sounds in English. We can represent all sounds phonemically with the same symbol, /k/. If we compare phonemic representations in general with phonetic representations, we see that phonemic representations eliminate information that is predictable. Consider the phonemic and phonetic representation of English : (1)

Phonemic transcription: Phonetic transcription:

/kiːp/ [k̟iːp]

The phonemic transcription of omits the predictable information about the advanced articulation of the sound, which is included in the phonetic transcription. In the following sections we will present the phonemes of German and discuss some of the issues involved in arriving at the set of phonemes presented here. In most cases, we can use minimal pairs as the basis for identifying phonemes. In some cases, though, additional considerations will play a role in our decisions. We will discuss some of the features that distinguish the phonemes of German, although we will not provide an exhaustive treatment of features. We will also discuss important allophonic variations. For example, we will take the position that [ç] and [x] are allophones of a single phoneme, /ç/, and we will describe the environments in which each of these allophones appears. Before we consider the consonant phonemes of German, we will say something about syllables—enough so that you will be able to understand the role they play in the distribution of the allophones of these consonants. The sounds in a word are grouped together into larger units called syllables. For example, the word ‘sea’ [zeː] has one syllable, ‘sailor’ [zeːman] contains two syllables, and ‘seamanlike’ [zeːmɛnɪʃ] is made up of three syllables. We indicate which sounds in a word belong to the same syllable by putting a period where the boundaries between syllables occur. In a word like ‘to believe’ /glaʊ̯.bən/, which is made up of two syllables, the syllable boundary shows that the segments /glaʊ̯/ belong to the first syllable and /bən/ belong to the second. Syllable boundaries also show us which segments occur at the beginning of a syllable and which occur at the end. In /glaʊ̯.bən/, for example, /b/ is at

46  Phonology the beginning of the second syllable and /n/ is at the end. We will include syllable boundaries in the phonemic representations of words in this chapter when syllables are relevant for the discussion. 3.3. The Consonant Phonemes of German Table 3.1 presents the consonant phonemes of German together with their allophones and an example of an environment in which each allophone can be found. We will begin our discussion by focusing on the phonemic contrasts that can be Table 3.1 The consonant phonemes of German. Phoneme

/p/ /b/ /t/ /d/ /k/ /g/

/m/ /n/ /ŋ/ /f/ /v/ /s/ /z/ /ʃ/ /ʒ/ /ç/ /ʀ/ /h/ /j/ /l/ /pf/ /ts/ /tʃ/ /dʒ/

Allophone

[p] [b] [p] [t] [d] [t] [k] [ g] [k] [ç] [m] [n] [ŋ] [f ] [v] [f ] [s] [z] [s] [ ʃ ] [ʒ] [ ʃ ] [x] [ç] [ʁ] [ɐ̯] [ɐ] [h] [j] [l] [pf ] [ts] [tʃ] [dʒ]

Example

/pas/ [pas] ‘passport’ /bas/ [bas] ‘bass’ /loːb/ [loːp] ‘praise’ /taɪ ̯ç/ [taɪ ̯ç] ‘pond’ /daɪ ̯ç/ [daɪ ̯ç] ‘dike’ /laɪ ̯d/ [laɪ ̯t] ‘sorrow’ /kaːbəl/ [kaːbəl] ‘cable’ /gaːbəl/ [ gaːbəl] ‘fork’ /laːg/ [laːk] ‘lay’ /ʀuːɪg/ [ʁuːɪç] ‘calm’ /hɛmən/ [hɛmən] ‘hindering’ /hɛnən/ [hɛnən] ‘hens’ /hɛŋən/ [hɛŋən] ‘hanging’ /faːzə/ [faːzə] ‘phase’ /vaːzə/ [vaːzə] ‘vase’ /bʀaːv/ [bʁaːf ] ‘well-behaved’ /ʀaɪ ̯sən/ [ʁaɪ ̯sən] ‘to tear’ /ʀaɪ ̯zən/ [ʁaɪ ̯zən] ‘to travel’ /glaːz/ [ glaːs] ‘glass’ /ʃaːl/ [ ʃaːl] ‘scarf ’ /ʀaːʒə/ [ʁaːʒə] ‘rage’ /beːʒ/ [beːʃ ] ‘beige’ /laçən/ [laxən] ‘to laugh’ /ʀaɪ ̯çən/ [ʁaɪ ̯çən] ‘to extend’ /ʀɪpə/ [ʁɪpə] ‘rib’ /meːʀ/ [meːɐ̯] ‘sea’ /bɪtəʀ/ [bɪtɐ] ‘bitter’ /hɔɪ ̯tə/ [hɔɪ ̯tə] ‘today’ /jʊŋə/ [jʊŋə] ‘boy’ /lɪpə/ [lɪpə] ‘lip’ /pfaɪ ̯l/ [pfaɪ ̯l] ‘arrow’ /tsaːm/ [tsaːm] ‘tame’ /tʃɛk/ [tʃɛk] ‘check’ /dʒɔb/ [dʒɔp] ‘job’

Phonology  47 found in German and justifying the phonemic status that we have given to the consonants in table 3.1. 3.3.1. Phonemic Contrasts Minimal pairs can be found for all the consonant phonemes that appear in ­table 3.1. For many of the phonemes, minimal pairs can be found among the examples in the table. If you look at the first phoneme in the table, /p/, you will see that the example for this phoneme, ‘passport’ [pas], forms a minimal pair with the example for /b/, ‘bass’ [bas]. The sounds [p] and [b] contrast at the very beginning of these two words. If you consider the nasal phonemes, you will see that the examples provided form a minimal triplet: ‘hindering’ [hɛmən], ‘hens’ [hɛnən], ‘hanging’ [hɛŋən]. We will not provide and discuss minimal pairs for each phoneme here but will focus our discussion instead on those phonemes that are less straightforward. The exercises at the end of the chapter will give you the opportunity to work with minimal pairs for all the phonemes of German. If you compare the consonant sounds of German (see table 1.1 in chapter 1) with the list of consonant phonemes in table 3.1, you will notice that although the glottal stop is one of the consonant sounds of German, it does not have the status of a phoneme. It is the case that we can find minimal pairs in which the glottal stop contrasts with other consonants. (2)

‘to remind’ [maːnən] ‘cloth’ [dɛkə] ‘chorus’ [koːɐ̯]

‘to suspect’ [ʔaːnən] ‘corner’ [ʔɛkə] ‘ear’ [ʔoːɐ̯]

However, the distribution of the glottal stop is predictable: it is found wordinitially before a vowel; within a compound before a word that begins with a vowel; and after a prefix if the following sound is a vowel. (3)

‘respect’ [ʔaxtʊŋ] ‘self-respect’ [zɛlpstʔaxtʊŋ] ‘to heed’ [bəʔaxtən]

The use of the glottal stop is a characteristic trait of German that gives it a certain choppiness not typically found in English. Because the glottal stop momentarily stops the flow of air, it has the effect of separating words that begin with a vowel from preceding words (and prefixes). In spite of the characteristic sound it gives the language, the glottal stop is not always pronounced. In quick, relaxed speech, it can be omitted. Because it is predictable and optional, the glottal stop is not considered a phoneme of German. Although the velar nasal, [ŋ], contrasts with the other nasals word-internally, as the minimal triplet , , and demonstrates (see

48  Phonology table 3.1), a number of linguists do not treat it as a phoneme of German (for example, Seiler 1962, Vennemann 1970, Lessen Kloeke 1982, Hall 1992, and Wiese 2000). Instead, they view it as derived from a sequence of two phonemes, either /ng/ or /nk/. We will not go into the details of this type of analysis (which involves providing an explanation for deriving the pronunciation for ‘to sing’ [zɪŋən], for example, from the phonemic representation /zɪngən/). We will point out, though, that this analysis explains why [ŋ] never occurs word-initially in German and why vowels before [ŋ] are always short. The sequence /ng/ is similar to /mb/ and /nd/, neither of which can occur word-initially in German, so it should not be surprising that /ng/ (pronounced as [ŋ]) cannot occur wordinitially. Only short vowels occur before /mb/ and /nd/, so it should not be unusual that only short vowels occur before /ng/: ‘snack’ [ʔɪmbɪs]; ‘to broadcast’ [zɛndən]; ‘cheek’ [vaŋə]. In spite of some of the explanatory advantages of assuming that [ŋ] is derived from a sequence of /ng/ or /nk/ (Benware 1986:47–48), we assume in this book that it has the status of a phoneme, /ŋ/. Because [s] and [z] contrast word-internally, as the minimal pair ‘to tear’ [ʁaɪ ̯sən] and ‘to travel’ [ʁaɪ ̯zən] demonstrates, we treat them as two different phonemes of German, /s/ and /z/. The distribution of these two sounds is limited in comparison to other consonant sounds in German, so you will have to be careful when looking for other minimal pairs to demonstrate that these two sounds contrast. [s] is never found word-initially before a vowel in a native German word; [z] is never found in word- or syllable-final position (because of final devoicing). One final issue in determining the inventory of consonant phonemes in ­German involves affricates. Some studies treat each of these sequences of stops and homorganic fricatives as a sequence of two separate phonemes (Moulton 1962, Heike 1972, Benware 1986, Kohler 1995); others treat them as a single phoneme (Hall 1992, Mangold 2005). We treat them as a single phoneme. We do not use special diacritics, however, to distinguish the monophonemic sequence /ts/ in a word like ‘time’ /tsaɪ ̯t/ [tsaɪ ̯t] from the sequence in a word like ‘(you) step’ /trɪtst/ [tʁɪtst], where the /t/ belongs to the stem and the /s/ belongs to the second person singular ending . 3.3.2. Allophonic Variation In this section we will talk about the allophonic variation illustrated in table 3.1. We will begin with the voiced obstruents (plosives and fricatives), each of which has a voiced and a voiceless allophone. If you compare the two allophonic variants for each of these phonemes in table 3.1, you should see a pattern in their distribution. The voiced variants appear syllable-initially; the voiceless variants appear in syllable-final position. The voiceless variants are the result of final devoicing, the process that requires all obstruents in German to be voiceless when

Phonology  49 they appear at the end of a word or syllable. We can illustrate how this works in the context of phonemes and allophones by considering the phoneme /b/ in two different forms of the word ‘praise’. (4)

‘praise (genitive singular)’ /loː.bəs/ [loː.bəs] ‘praise’ /loːb/ [loːp]

Both forms of the word have the voiced bilabial plosive phoneme, /b/. In the genitive form, /b/ is in syllable-initial position (it is at the beginning of the second syllable in the word), so it is pronounced as a voiced bilabial plosive. In the other singular forms of the word, the /b/ is in syllable-final position (it is at the end of the word, and the word consists of only one syllable), so it is subject to final devoicing and is pronounced as a voiceless bilabial plosive, [p]. Keep in mind that final devoicing affects voiced fricatives as well as plosives. The voiced labiodental fricative phoneme, /v/, for example, is pronounced as the voiceless labiodental fricative, [f ], when it occurs in syllable-final position. The inflected and uninflected forms of the adjective ‘well-behaved’ demonstrate this. (5)

‘well-behaved’ /bʀaː.və/ [bʁaː.və] ‘well-behaved’ /bʀaːv/ [bʁaːf ]

When has the ending , for example, /v/ occurs syllable-initially and is voiced. When the adjective does not have an ending, /v/ is in syllable-final position and therefore voiceless. The voiced alveolar fricative phoneme, /z/, is affected by final devoicing in the same way, as the plural and singular forms of the word ‘track’ demonstrate. (6)

‘tracks’ /glaɪ ̯.zə/ [ glaɪ ̯.zə] ‘track’ /glaɪ ̯z/ [ glaɪ ̯s]

Both forms of the word have the phoneme /z/. In the plural form, /z/ is in syllable-initial position and is therefore pronounced as a voiced alveolar fricative, [z]. In the singular form, because /z/ is in syllable-final position, it undergoes devoicing and is pronounced [s]. In the exercises at the end of this chapter, you will be given phonemic representations of words and asked to transcribe them phonetically. When a phonemic transcription contains a voiced plosive phoneme (/b/, /d/, or /g/) or a voiced fricative phoneme (/v/, /z/, or /ʒ/), you will need to pay attention to the position of the phoneme when you transcribe it phonetically. If it occurs at the beginning of a syllable, you will transcribe it as a voiced plosive or fricative, but if it occurs at the end of a syllable, you will need to transcribe it phonetically as a voiceless plosive or fricative, that is, as [p], [t], [k], [f ], [s], or [ ʃ ]. Another important point that we need to make is that all the obstruents at the end of a syllable

50  Phonology must be ­transcribed phonetically as voiceless. Even if an obstruent is not in absolute syllable-final position, it will be voiceless as long as it is one of any of the consonants at the end of a syllable. In a word like ‘(you) live’ /leːbst/, for example, which consists of a single syllable, /b/ is realized as a voiceless plosive, [p], because it is one of the consonants at the end of the syllable: [leːpst]. The /z/ in ‘travels’ /ʀaɪ ̯zt/ is realized as [s] for the same reason: [ʁaɪ ̯st]. The phoneme /g/ is a bit more complex than the other two voiced plosive phonemes in German, /b/ and /d/, because it has not just two, but three allophones. In addition to its voiced and voiceless plosive allophones, [ g] and [k], it has the allophone [ç], a fricative. This allophone is found when /g/ is in the ending /-ɪg/ and the ending is not followed by a vowel. In an adjective like ‘holy’, for example, /g/ is pronounced as [ç] because it occurs in the suffix /ɪg/ at the very end of the word: /haɪ ̯.lɪg/ [haɪ ̯.lɪç]. In the verb form ‘keeps holy’, /g/ is also pronounced [ç] because the suffix /ɪg/ is followed by a consonant: /haɪ ̯.lɪgt/ [haɪ ̯.lɪçt]. If the suffix /ɪg/ is followed by a vowel, as in , an inflected form of the adjective , then /g/ is pronounced as [ g] because it is at the beginning of the syllable: /haɪ ̯.lɪ.gə/ [haɪ ̯.lɪ.gə]. When you have a phonemic representation that contains the phoneme /g/, you will first need to determine whether it occurs in the ending /-ɪg/. If it does, then /g/ will be realized (and transcribed) phonetically as [ç] unless it occurs before a vowel, in which case it will be realized as [ g]. If /g/ does not occur in the ending /-ɪg/, then it will be realized as [ g] if it occurs at the beginning of a syllable or as [k] if it occurs at the end of a syllable. We have already pointed out that the phoneme /ç/ has two allophones, [x] and [ç]. The simplest way to describe the distribution of these allophones is to say where [x] appears and then say that [ç] appears “elsewhere”: [x] appears after /aː/, /a/, /aʊ̯/, and the back vowels (/uː ʊ oː ɔ/); [ç] appears elsewhere. Notice that [x] and [ç] are in complementary distribution; these sounds never occur in the same environment. Given a phonemic representation like /buːç/, for example, you will be able to predict that /ç/ will be realized as [x] because it follows a back vowel, /uː/. In /mɪç/, on the other hand, it follows a front vowel, so it will be realized as [ç]. The final phoneme with more than one realization is the phoneme that is represented in German orthography with the letter . Although the voiced uvular fricative, [ʁ], is the dominant consonantal pronunciation of and therefore the pronunciation we use here, we do not use it to represent the phoneme. One of our reasons for doing this is that the phoneme does not behave like other voiced fricative phonemes. In particular, it does not undergo final devoicing, as we will see shortly. We use instead the symbol for the uvular trill, /ʀ/. When /ʀ/ appears before a vowel in the same syllable, it is pronounced as [ʁ]. Both instances of /ʀ/ in a word like ‘sad’, for example, have a consonantal pronunciation: /tʀaʊ̯.ʀɪg/ [tʁaʊ̯.ʁɪç]. When /ʀ/ follows a vowel in the

Phonology  51 same ­syllable, it has a vocalic pronunciation; it is pronounced as nonsyllabic dark schwa, [ɐ̯]. In the monosyllabic word ‘sea’, for example, /ʀ/ follows the vowel and is therefore pronounced as a nonsyllabic dark schwa: /meːʀ/ [meːɐ̯]. If the vowel that /ʀ/ follows is schwa, /ə/, then the combination /əʀ/ is pronounced as dark schwa, [ɐ]. In a word like ‘louder’, /ʀ/ follows a schwa in the same syllable, so the /əʀ/ sequence is pronounced as dark schwa: /laʊ̯.təʀ/ [laʊ̯.tɐ]. Notice that not all sequences of /əʀ/ are pronounced as dark schwa. If a word like has an adjective ending like , /ʀ/ will have a consonantal pronunciation (and the preceding /ə/ will be pronounced as [ə]) because /ʀ/ is not in the same syllable as the preceding /ə/. It is in the following syllable and precedes the vowel in this syllable: /laʊ̯.tə.ʀə/ [laʊ̯.tə.ʁə]. When you are faced with a phonemic transcription with the phoneme /ʀ/ and want to determine how to transcribe it phonetically, you will need to consider the environment in which it appears: Does /ʀ/ precede or follow the vowel in its syllable? If it follows a vowel in the same syllable, is this vowel schwa? 3.4. The Vowel Phonemes of German Table 3.2 presents the vowel phonemes of German together with their allophones and an example of an environment in which each allophone can be found. The discussion in the following section will focus on the important phonemic contrasts among the vowels and the justification for recognizing the phonemes in table 3.2 as the vowel phonemes of German. 3.4.1. Phonemic Contrasts The minimal pair ‘lame’ [laːm] and ‘lamb’ [lam] demonstrates that [aː] and [a] are distinctive sounds in German and can be given the status of phonemes. The single feature that distinguishes these two phonemes is length: /aː/ is [+long]; /a/ is [−long]. On the basis of these two phonemes, we can say that the feature [long] is a distinctive feature of German. It is the single feature that distinguishes these two phonemes. The feature of tenseness does not distinguish these two phonemes, since neither is marked for tenseness. There are an additional six pairs of vowel phonemes that differ in length, one of which is the pair /iː ɪ/. On the basis of a minimal pair like ‘to offer’ [biːtən] and ‘to ask’ [bɪtən], we can give both of these vowels the status of phonemes in German. These two phonemes do not just differ in length, however—they also differ in tenseness: /iː/ is [+tense]; /ɪ/ is [−tense]. This is also true of the pairs /yː ʏ/, /uː ʊ/, /eː ɛ/, /øː œ/, and /oː ɔ/. If a vowel in German is marked for tenseness (if it is [+tense] or [−tense]), then we can predict whether it is long or short. Vowels that are [+tense] are also [+long]; those that are [−tense] are [−long]. Do we need to specify that the tense vowel phonemes of German are long? Could we just eliminate this information, since it is predictable? No—not without

52  Phonology Table 3.2 The vowel phonemes of German. Phoneme /iː/

/ɪ/ /yː/ /ʏ/ /uː/ /ʊ/ /eː/ /ɛ/ /øː/ /œ/ /ə/ /oː/ /ɔ/ /aː/ /a/ /aɪ ̯/ /aʊ̯/ /ɔɪ ̯/

Allophone

[iː] [i] [ɪ] [ yː] [ y] [ʏ] [uː] [u] [ʊ] [eː] [e] [ɛ] [øː] [ø] [œ] [ə] [oː] [o] [ɔ] [aː] [a] [a] [aɪ ̯] [aʊ̯] [ɔɪ ̯]

Example

/muːˈziːk/ [muˈziːk] ‘music’ /muːziːˈkaːlɪʃ/ [muziˈkaːlɪʃ] ‘musical’ /ˈmɪtə/ [ˈmɪtə] ‘middle’ /ˈfyːziːkəʀ/ [ˈfyːzikɐ] ‘physicist’ /fyːˈziːk/ [fyˈziːk] ‘physics’ /ˈmʏsən/ [ˈmʏsən] ‘to have to’ /ˈjuːbəl/ [ˈjuːbəl] ‘jubilation’ /juːbiːˈliːʀən/ [jubiˈliːʁən] ‘to jubilate’ /ˈmʊtəʀ/ [ˈmʊtɐ] ‘mother’ /ˈleːbən/ [ˈleːbən] ‘to live’ /leːˈbɛndɪg/ [leˈbɛndɪç] ‘lively’ /nɛt/ [nɛt] ‘nice’ /ˈøːkoː/ [ˈʔøːko] ‘Green’ /øːkoːloːˈgiː/ [ʔøkoloˈgiː] ‘ecology’ /ˈkœnən/ [ˈkœnən] ‘to be able to’ /ˈbɪtə/ [ˈbɪtə] ‘please’ /ˈpʀoːbə/ [ˈpʁoːbə] ‘test’ /pʀoːˈbiːʀən/ [pʁoˈbiːʁən] ‘to try’ /ˈvɔlə/ [ˈvɔlə] ‘wool’ /ʀeːˈaːl/ [ʁeˈaːl] ‘real’ /ʀeːaːliːˈteːt/ [ʁealiˈteːt] ‘reality’ /ˈpanə/ [ˈpanə] ‘breakdown’ /daɪ ̯n/ [daɪ ̯n] ‘your’ /baʊ̯m/ [baʊ̯m] ‘tree’ /nɔɪ ̯/ [nɔɪ ̯] ‘new’

complicating our analysis of allophonic variation in the long vowel phonemes. If you consider table 3.2, you will see that each of the long vowel phonemes in German has a short as well as a long allophone. The phoneme /iː/, for example, is realized as both [iː] and [i]. We will say more about the distribution of these allophones in the following section. For now, we just want to point out that we can account for the distribution of these allophones in a very straightforward way if we assume that the tense vowel phonemes in German are marked for length as well as tenseness. Another feature that plays a role in distinguishing the vowel phonemes of German is [rounded]. This feature distinguishes the phoneme /yː/, for example, from the phoneme /iː/; it is the single feature that distinguishes these two phonemes: /yː/ is [+rounded], /iː/ is [−rounded]. There are a total of four pairs of vowels in German that differ only in rounding. We will not discuss these further here, but will save the details for the exercises at the end of this chapter.

Phonology  53 There are four more features that play a role in distinguishing the vowel phonemes in German. Two account for differences in tongue height: [high], [low]. Two account for differences in tongue position: [front], [back]. We will demonstrate how these pairs of features work by focusing on how the features [front] and [back] can be used to distinguish three different tongue positions: front, central, and back. A front vowel like /ɛ/, for example, is [+front] and [−back]; a back vowel like /ɔ/ is [−front] and [+back]. A central vowel like /ə/, on the other hand, is neither front nor back, so we can treat it as [−front] and [−back]. (The only other combination of these two features, [+front] and [+back], is physically impossible.) Notice that although only one feature distinguishes /ə/ from /ɛ/ ([front]), there are two features that distinguish /ə/ from /ɔ/ ([back] and [rounded]). We will leave further discussion of features for the exercises at the end of this chapter so that you will have the opportunity to determine for yourself the ways in which these features distinguish the different vowel phonemes of German. One phoneme of German deserves special discussion because of its uniqueness and the attention that it has attracted in the study of German phonology: /ə/. Schwa differs from all the other vowel phonemes of German in that it never occurs in a stressed syllable. Although it is not easy to find minimal pairs in which schwa contrasts with other vowels in unstressed syllables, such pairs do exist, as the words ‘rose’ [ʁoːzə] and ‘pink’ [ʁoːza] demonstrate. Some studies do not consider schwa to be a phoneme of German because its presence can often be predicted, much like the presence of the glottal stop. This treatment of schwa is not uncontroversial, however. For further discussion, you can consult studies such as Giegerich 1987, Wiese 1988, Hall 1992, and Becker 1998. To keep our treatment of German phonology as simple and straightforward as possible, we give schwa the status of a phoneme. Although dark schwa, [ɐ], contrasts with schwa, as the minimal pair ‘bitter’ [bɪtɐ] and ‘please’ [bɪtə] demonstrates, we do not treat dark schwa as a phoneme of German. If we consider different forms of words that contain the sound [ɐ], we see that [ɐ] alternates with [əʁ]. (7)

[ɐ] – [əʁ] ‘clean’ [zaʊ̯bɐ] ‘more beautiful’ [ ʃøːnɐ] ‘to wander’ [vandɐn]

‘clean’ [zaʊ̯bəʁə] ‘more beautiful’ [ ʃøːnəʁə] ‘(I) wander’ [vandəʁə]

In the stem , for example, the sequence can be pronounced as [ɐ] or as [əʁ] (depending on the syllable structure of the word in which the stem appears). This also holds for in the stems and . If we represent the sequence in all of these words as the sequence of phonemes /əʀ/, then we can predict when it is pronounced as [ɐ] and when it is pronounced

54  Phonology as [əʁ]. Recall from the discussion of the /ʀ/ phoneme that the sequence /əʀ/ is pronounced as [ɐ] if /ə/ and /ʀ/ belong to the same syllable; otherwise it is pronounced as [əʁ]. In the infinitive , /ə/ and /ʀ/ belong to the same syllable (and /ʀ/ follows /ə/), so /əʀ/ is pronounced as dark schwa: /van.dəʀn/ [van.dɐn]. In the first person singular present tense form of the verb, ‘(I) wander’, the /ə/ and /ʀ/ in the sequence /əʀ/ belong to two different syllables (and /ʀ/ precedes the vowel in its syllable), so the sequence is pronounced [əʁ]: /van.də.ʀə/ [van.də.ʁə]. Although arguments can be made for treating diphthongs as a sequence of two vowel phonemes, we follow common practice and treat the diphthongs in German as single phonemes. They contrast with single vowels, both tense and lax, as the following minimal pairs demonstrate. (8)

‘to ride’ [ʁaɪ ̯tən] ‘lazy’ [faʊ̯l] ‘pack’ [mɔɪ ̯tə]

‘to redden’ [ʁøːtən] ‘much’ [fiːl] ‘rent’ [miːtə]

‘to save’ [ʁɛtən] ‘full’ [fɔl] ‘moth’ [mɔtə]

For a discussion of the arguments for and against treating the German diphthongs as single phonemes, see Becker 1998. 3.4.2. Allophonic Variation Before we discuss allophonic variation in German vowels, we need to briefly describe what it means for a syllable to be stressed. Put simply, a stressed syllable is one that is perceived as more prominent than other syllables. Prominence is a complex phenomenon and can involve factors such as duration, loudness, and pitch. In German, the most reliable correlate of prominence is duration: stressed syllables are longer than unstressed syllables. We use the diacritic [ˈ] to indicate that the following syllable receives primary stress—it is the most prominent syllable in the word. In a word like ‘to receive’, for example, the second syllable in the word receives primary stress: /bəˈkɔmən/. All the phonemes in German that are long phonemically have short as well as long allophones. This includes the tense vowels, /iː yː uː eː øː oː/, and /aː/ (which is long but neither tense nor lax). The long and short allophones of each of these phonemes are in complementary distribution: the long allophones are found in stressed syllables, the short ones are found in unstressed syllables. The root , for example, has a long tense sound phonemically, /oː/. When this root appears in the noun ‘test’, /oː/ is realized phonetically as long because the syllable in which it occurs is stressed: /ˈpʀoːbə/ [ˈpʁoːbə]. When the root occurs in the verb , the /oː/ is still a tense vowel, but it is short (it does not last as long as the sound in ) because it is in a syllable that is not stressed: /pʀoːˈbiːʀən/ [pʁoˈbiːʁən].

Phonology  55

Why have phonemic transcriptions? Why not just phonetic transcriptions? 1. We can view phonemic transcriptions as the ideal orthography for a language, because each symbol represents a distinct sound of the language. Phonemic transcriptions give us insight into what constitutes a distinct sound to a native speaker of the language. For a native speaker of German, the /b/ in /leːbən/ is the same sound as the /b/ in /leːbt/; /leːbən/ and /leːbt/ are two forms of the same word, and both have the sound /b/. Native speakers of German do not realize that they are pronouncing the /b/ in /leːbt/ as [p]; they do this automatically. 2. Phonemic transcriptions are simpler (less detailed) than phonetic transcriptions, since all of the predictable information has been filtered out. We do not need to use the symbols [ɐ] or [ɐ̯] in our phonemic representation of German; we can just use /ʀ/ and /ə/, which we already have, since we can predict when /ʀ/ will be pronounced [ɐ̯] and when /əʀ/ will be pronounced [ɐ]. 3. Phonetic transcriptions tell us how words are pronounced. Phonetic transcriptions can be broad (not highly detailed) or narrow (with all kinds of phonetic detail). The phonetic transcriptions we use in this book are typically broad. They give us the basic details of how words in German are pronounced and they provide the standard pronunciation of words. What phonetic transcriptions do not tell us is how sounds function or pattern in a language. For example, the phonetic transcriptions [dɪç] and [dɔx], for two words in German, tell us that German has the sounds [ç] and [x], but these transcriptions do not tell us that these function as variants of what Germans consider to be one sound. The bottom line: Phonemic transcriptions are not better than phonetic transcriptions or vice versa. They simply provide different kinds of information.

We will discuss word stress in German in more detail in chapter 4. In the exercises in this chapter, we will indicate stress in the phonemic representation of all words of more than one syllable when stress is relevant for the exercise. For example, if an exercise requires you to determine how long vowels are realized phonetically, the necessary information regarding stress will be provided. 3.5. Advanced Topics 3.5.1. Nasalized Vowels In chapter 1 we mentioned nasalized vowels that occur in words like ‘restaurant’ [ʁɛstoʁãː], ‘perfume’ [paɐ̯fœ̃ ː], and ‘balcony’ [balkõː] (nasalization is indicated with a tilde, [˜]). Although the Duden

56  Phonology ­ ronunciation dictionary treats nasalized vowels as phonemes of German (Manp gold 2005:35), most accounts do not consider them part of the German vowel system. Remember that nasalized vowels are typically found in loanwords from French. However, just because a sound occurs in a loanword does not automatically cause us to exclude it from the sound system of a language. Wiese (2000:12) excludes nasalized vowels from the phoneme system of German because of the tendency of speakers to change them by replacing them with oral vowels followed by a nasal consonant. Alternative pronunciations of , for example, are [balkɔŋ] and [balkoːn]. Wiese claims that nasalized vowels “rarely occur in normal speech.” On the other hand, if speakers of a language readily accept a sound in a borrowed word, without any noticeable tendency to change the sound in some way, Wiese argues that the sound belongs to the phonological system of this language. According to Wiese, an example of a sound in German that has this characteristic is /ʒ/, which was introduced in German primarily through French loanwords. Wiese (2000:12) says that there “seems to be no tendency to assimilate /ʒ/ to the system of more native sounds,” and thus includes it in the inventory of German phonemes. However, the Duden pronunciation dictionary notes that /ʒ/ is replaced by its voiceless counterpart, [ ʃ ], in the colloquial (casual) speech of some speakers. For example, some speakers pronounce ‘disgrace’ as [blamaːʃə] rather than [blamaːʒə] (Mangold 2005:66). We follow Wiese and include /ʒ/ as a phoneme of German, but we do not include nasalized vowels. The substitution of [ ʃ ] for [ʒ] is not considered a standard pronunciation, whereas the substitution of a vowel and nasal consonant for a nasalized vowel has become standard for many words. Some words only have a pronunciation with a vowel followed by a nasal consonant: ‘garrison’ [ gaɐ̯nizoːn]; ‘battalion’ [bataljoːn]. Others have nasalized vowels as well as vowels followed by a nasal consonant: ‘balloon’ [balõː], [balɔŋ], [baloːn]; ‘drawing room’ [zalõː], [zalɔŋ], [zaloːn]. Others appear only with nasalized vowels in the Duden pronunciation dictionary, but very likely have colloquial pronunciations with vowels followed by a nasal consonant. One such word is ‘restaurant’ [ʁɛstoʁãː], which Kohler (1995:173) cites as being pronounced with [aŋ] in northern Germany. 3.5.2. The Allophones of /ʀ/ Although the Duden pronunciation dictionary (Mangold 2005) uses the alveolar trill, [r], to transcribe all instances of consonantal , we use the voiced uvular fricative, [ʁ], since it is the dominant pronunciation of consonantal in German. Another difference between Duden’s treatment of and the approach we follow here involves the pronunciation of after short vowels. Duden transcribes a consonantal after a short vowel; we transcribe a nonsyllabic dark schwa, [ɐ̯].

Phonology  57 If you look up the word ‘there’, for example, which we transcribe as [dɔɐ̯t], you will see that Duden transcribes it as [dɔrt]. Duden notes, however, that both consonantal and vocalic are possible after short vowels (Mangold 2005:54). We transcribe a vocalic in this position because this simplifies our description of the distribution of the allophones of /ʀ/—[ʁ] occurs in prevocalic position within a syllable, while [ɐ̯] occurs in postvocalic position. Exercises 3.1. Minimal pairs. The words ‘passport’ and ‘bass’ form a minimal pair and show that [p] and [b] contrast in German and that they can thus be considered phonemes. Demonstrate that the following sets of sounds contrast in German by finding a minimal pair for each set. a. [p] : [t] b. [d] : [n] c. [ ʃ ] : [z] d. [m] : [b] e. [ j] : [h] f. [t] : [k] g. [f ] : [s] h. [h] : [ʁ] i. [v] : [z] j. [l] : [n] 3.2. Affricates. For each of the affricates below, find a minimal pair in which the affricate contrasts with a single consonant sound. a. [pf ] b. [ts] c. [tʃ] d. [dʒ] 3.3. Final devoicing. Each of the following words is accompanied by a phonemic transcription and an incomplete phonetic transcription. Indicate how the voiced obstruent phonemes in these words are pronounced by completing the phonetic transcriptions. Remember that all obstruents at the end of a syllable must be transcribed phonetically as voiceless. a. /ʃaː.dən/ [ ʃaː ən] ‘to damage’ b. /boː.dən/ [ oː ən] ‘ground’ c. /baːd/ [ aː ] ‘bath’ d. /fʀaːgt/ [fʁaː t] ‘asks’ e. /bə.gaːbt/ [ ə aː t] ‘talented’ f. /zaft/ [ aft] ‘juice’ g. /løː.zən/ [løː ən] ‘to solve’

58  Phonology h. /løːzt/ i. /zeː.lɛk.tiːv/ j. /ak.tiː.viː.ʀən/ k. /ʒeː.niː/ l. /glaːz.ʃaɪ ̯.bə/ m. /bə.tsaʊ̯.bəʀnd/ n. /zʏn.də/ o. /aː.bənds/

[løː t] [ elɛktiː ] [ʔakti iːʁən] [ eniː] [ laː ʃ aɪ ̯ ə] [ ətsaʊ̯ ɐn ] [ ʏn ə] [ʔaː ən s]

‘solves’ ‘selective’ ‘to activate’ ‘genius’ ‘pane of glass’ ‘enchanting’ ‘sin’ ‘in the evening’

3.4. [ g], [k], or [ç]? Indicate how /g/ is realized phonetically in the following words. a. /gɪf.tɪg/ [ ɪftɪ ] ‘poisonous’ b. /gə.aɪ ̯.nɪgt/ [ əʔaɪ ̯nɪ t] ‘united’ c. /flaɪ ̯.sɪ.gəʀ/ [flaɪ ̯sɪ ɐ] ‘more diligent’ d. /meː.lɪg/ [meːlɪ ] ‘mealy’ e. /taɪ ̯g/ [taɪ ̯ ] ‘dough’ f. /fɔl.gə/ [fɔl ə] ‘consequence’ g. /bə.glaʊ̯.bɪgt/ [bə laʊ̯bɪ t] ‘certified’ h. /gaɪ ̯gst/ [ aɪ ̯ st] ‘(you) fiddle’ i. /gʏl.tɪg/ [ ʏltɪ ] ‘valid’ j. /jaːgd/ [ jaː t] ‘hunt’ 3.5. Which allophone of /ç/ is realized in the following words? [x] or [ç]? a. [ ] /mɪç/ ‘me’ b. [ ] /vɛlç/ ‘which’ c. [ ] /ˈʀaʊ̯.çən/ ‘to smoke’ d. [ ] /çeːˈmiː/ ‘chemistry’ e. [ ] /ˈmœç.tə/ ‘would like to’ f. [ ] /kɔç/ ‘cook’ g. [ ] /ˈbuː.çən/ ‘to book’ h. [ ] /ˈbɪs.çən/ ‘bit’ i. [ ] /dʊʀç/ ‘through’ j. [ ] /faç/ ‘subject’ 3.6. The realization of /ʀ/ and /ə/. Fill in the missing phonetic details. a. /ʀaɪ ̯.bən/ [ aɪ ̯b n] ‘to rub’ b. /ʀɔɪ ̯.bəʀ/ [ ɔɪ ̯b ] ‘robber’ c. /dɔʀt/ [dɔ t] ‘there’ d. /dʀaː.maː.tiː.kəʀ/ [d amaːtik ] ‘dramatist’ e. /fɔɪ ̯.əʀ/ [fɔɪ ̯ ] ‘fire’ f. /fʀiː.ziː.ʀən/ [f iziː n] ‘to do someone’s hair’ g. /ʀaɪ ̯.fə.ʀə/ ] [ aɪ ̯f ‘more mature’ h. /ʃtɔl.pəʀn/ [ ʃtɔlp n] ‘to stumble’

Phonology  59 i. / ʃtɔl.pəʀ.tə/ j. /ʀeː.duː.tsiːʀt/

[ ʃtɔlp t ] [ edutsiː t]

‘stumbled’ ‘reduced’

3.7. /ʔ/. Underline the vowels in the following words that are preceded by a glottal stop in careful speech. a. ‘breath’ b. ‘clocks’ c. ‘photograph album’ d. ‘to end’ e. ‘open’ f. ‘to remember’ g. ‘quiet’ h. ‘wristwatch’ i. ‘abnormal’ j. ‘evening’ 3.8. Minimal pairs. Demonstrate that the following sets of vowels contrast in German by finding a minimal pair for each set. a. [iː] : [ yː] b. [ɪ] : [iː] c. [eː] : [oː] d. [uː] : [ yː] e. [ʊ] : [ʏ] f. [eː] : [ɛ] g. [œ] : [ɔ] h. [aɪ ̯] : [aː] i. [aʊ̯] : [oː] j. [ɔɪ ̯] : [aɪ ̯] 3.9. Tongue height. In chapter 1 we identified three different tongue heights necessary to characterize the vowels of German: high, mid, low. We can also do this with two binary features, [high] and [low] (binary features are marked either + or  −). Use these two features to distinguish the phonemes /iː/, /eː/, and /aː/. (Note: /iː/ and /eː/ are both front vowels, whereas /aː/ is a central vowel, so /iː/ and /eː/ differ from /aː/ in tongue position as well as tongue height.) 3.10. Distinctive features. Each of the following pairs of phonemes differs by one feature. Name the single feature that distinguishes them: [long], [rounded], [high], [low], [front], [back]. a. /iː/ : /yː/ b. /iː/ : /eː/ c. /aː/ : /a/ d. /uː/ : /oː/

60  Phonology e. /œ/ : /ɛ/ f. /ɛ/ : /ə/ g. /ʏ/ : /ɪ/ h. /ɛ/ : /ɪ/ i. /eː/ : /øː/ j. /œ/ : /ʏ/ 3.11. Vowel shortening. Each of the following words is accompanied by a phonemic transcription and an incomplete phonetic transcription. Indicate how the long vowel phonemes in these words are pronounced by completing the phonetic transcriptions. Remember that if a long vowel is not stressed, it will be realized as a short vowel phonetically, but it will not change in tenseness. a. /deːˈfɛkt/ [de ˈfɛkt] ‘defective’ b. /iːˈdeː/ [ʔi ˈde ] ‘idea’ c. /meːloːˈdiː/ [me lo ˈdi ] ‘melody’ d. /deːtɛkˈtiːv/ [de tɛkˈti f ] ‘detective’ e. /toːˈmaːtə/ [to ˈma tə] ‘tomato’ f. /ʃtuːˈdiːʀən/ [ ʃtu ˈdi ʁən] ‘to study’ g. /øːkoːˈnoːmɪʃ/ [ʔø ko ˈno mɪʃ ] ‘economical’ h. /bəˈgeːən/ [bəˈge ən] ‘to commit’ i. /myːtoːloːˈgiː/ [my to lo ˈgi ] ‘mythology’ j. /iːdeːaːˈlɪsmʊs/ [ʔi de a ˈlɪsmʊs] ‘idealism’ 3.12. Transcribe the words in exercise 3.5 phonetically. 3.13. Transcribe the following words (phrases) phonetically. Remember to transcribe glottal stops in the environments in which they occur in careful speech. a. /heːʀ/ ‘army’ b. /ˈʊn.təʀ.ʀɪçt/ ‘instruction’ c. /gə.ˈzaːgt/ ‘said’ d. /ˈliːb.lɪç/ ‘lovely’ e. /ˈʀaɪ ̯.zə.lʊst/ ‘wanderlust’ f. /ˈʀaɪ ̯z.tə/ ‘traveled’ g. /ˈoː.nə ˈɛn.də/ ‘without end’ h. /am ˈflaɪ ̯.sɪg.stən/ ‘most industrious’ i. /ˈveː.nɪ.gəʀ/ ‘less’ j. /̍ʃtaʊ̯b.zaʊ̯.gəʀ/ ‘vacuum cleaner’ k. /foː.ˈneː.tɪk/ ‘phonetics’ l. /ˈmoːnd.ʃaɪ ̯n/ ‘moonlight’ m. /ɛʀ.ˈfɔʀ.dəʀn/ ‘to require’ n. /ʃoː.koː.ˈlaː.də/ ‘chocolate’ o. /ˈaɪ ̯ç.hœʀn.çən/ ‘squirrel’

Phonology  61 p. /ˈgʀʊnd.ʀɛçt/ q. /ɪn.tʀoː.vɛʀ.ˈtiːʀt/ r. /ˈgɛld ʊnd ˈguːt/ s. /ˈmœn.çə/ t. /aɪ ̯n ˈʀaɪ ̯.çəʀ ˈhɛʀ/

‘fundamental right’ ‘introverted’ ‘wealth and possessions’ ‘monks’ ‘a rich gentleman’

This page intentionally left blank

Chapter 4

Syllables and Word Stress

4.1. Introduction Do you remember counting the number of beats in words when you were a child? Did you know that each beat in a word corresponds to a syllable? The fact that preschool-aged children are aware of and can make use of beats provides evidence that speakers are aware of syllables at a very early age. Young children are also able to play games that both rely on and manipulate syllable structure. They can do things like rhyme words and exchange vowels in words, as in the German children’s song “Drei Chinesen mit dem Kontrabass” (and the English song “Apples and Bananas”). Syllables play an important role in language itself, not just in the ways we put language to use. In German, for example, syllables are important for a number of phonological processes, including final devoicing, the insertion of glottal stops, and the realization of . In this chapter you will learn how syllables are built and how syllable structure is relevant for determining the pronunciation of German words. We will also talk about word stress, and you will learn how to use the structure of words to determine which syllable in a word bears primary stress. 4.2 . Sonority It is often straightforward to determine the number of syllables in a word. At a minimum, a German syllable must contain a vowel, like in the interjection ‘oh’ [ʔoː]. German words like ‘the’ or ‘a’ are made up of only one syllable, and relatively long words like ‘philosophic’ and ‘coffee cup’ both have four syllables. In this section, we will provide you with background into one of the basic principles underlying the notion of the syllable: sonority. Syllables are not simply composed of segments in a random order. Rather, they are built according to a principle called sonority, or relative loudness of sounds. This general principle allows us to categorize segments into two basic classes: sonorants, which are vowel-like sounds including vowels, /l/, /ʀ/, and nasals; and obstruents, which include plosives, fricatives, and affricates. Syllables are built up according to what is called the sonority hierarchy, which can be found in (1), based on Wiese 2011:72. The symbol “>” stands for “is/are more sonorous than.” 63

64  Syllables and Word Stress (1)



Sonority hierarchy for German nonhigh vowels > high vowels > /ʀ/ > /l/ > nasals > obstruents

If you look at the sonority hierarchy in (1), you will notice that a segment’s sonority corresponds to the relative freedom of airflow in the vocal tract during its articulation. Your mouth is open wider when you produce a low vowel like /a/ than it is when you produce a high, front vowel like /i/. Similarly, air can escape more freely when you produce a lateral approximant like /l/ than when you produce a plosive like /p/. Syllables in German are built according to the sonority principle in (2). (2)

Sonority principle Build syllables in such a way that the most sonorous (loudest) segments are in the center and less sonorous segments are at the edges.

If you consider a syllable like that in (3), you will see that the relative loudness of the syllable increases at the center of the syllable and decreases at its edges. (3)

Relative sonority of segments in the German word /traŋk/ ‘drank’



/t

ʀ

a

ŋ k/

The least sonorous segments, /t/ and /k/, both of which are obstruents, are at the edges of the syllable. Because /ʀ/ and /ŋ/ are more vowel-like, they appear closer to the center of the syllable, which contains the vowel /a/. Exercises at the end of the chapter will provide you with practice in counting syllables and applying the sonority principle. We will refer to the sonority hierarchy in (1) when we discuss how German syllables are built and how we determine the location of the edges of syllables. 4.3. Phonotactics The sonority principle provides us with a basic guideline for determining how syllables are built. It allows us to create new syllables in a principled manner. For example, we can swap out one obstruent from the syllable in (3) with another. We can also switch the vowel. The result is different syllables with the same basic sonority shape, as in (4). (4)

a. ‘sick’, ‘Frank’ b. ‘drink!’, ‘potion’

Syllables and Word Stress  65 In (4a) we have replaced the initial obstruent /t/ with other obstruents, /k/ and /f/. The words in (4b) are variants of (3) that contain different vowels. The sonority principle suffices when we are dealing with monosyllabic words. When we wish to syllabify longer words, though, we need to determine the location of syllable boundaries. This is especially important in German, since processes like final obstruent devoicing and the realization of critically depend on the location of syllable boundaries. In this section we will outline a number of general principles that can help you determine how to locate the boundaries between syllables. Phonotactics, the branch of phonology that deals with determining the sequences of segments that are permissible in a language, provides us with a set of constraints for determining—and thereby also limiting—which syllables are allowed. Unlike the sonority principle, which applies across languages, phonotactic constraints are language specific. Phonotactic constraints are especially helpful for determining which consonant phonemes may co-occur within a single syllable. For example, although German does not allow a series of two fricatives like /vz/ at the beginning of a syllable, Russian does. On the other hand, German allows /kn/ to occur at the beginning of a syllable, as in (5a), but English does not, even though this is an allowable sequence according to the sonority principle. Remember that although the sequence appears orthographically in English, this combination represents /n/, as in /naɪ ̯t/. The phonotactic rules of English disallow the sequence /kn/ at the beginning of a syllable. The sequence /kn/ does not appear at the end of a syllable in either English or German, since this is a violation of the sonority principle, as shown in (5b), where the asterisk indicates that the sequence of segments following it is not allowed. (5) Relative sonority of /kn/ in German at the beginning and end of syllables a.

/k n

a

l

t/

b.

*/∫

l

a

k

n/

66  Syllables and Word Stress The basic phonotactic rules of German consonant clusters tell us which consonant clusters are allowed at the beginning of German syllables (see table 4.1) and at the end (see table 4.2). In both tables, the segments in the column on the left are the potential first segments of the clusters, and those in the row on the top are the potential second segments of the clusters. Clusters that appear with “(+)” are those that occur only rarely in German, most often in loanwords. According to the constraints in table 4.1, it is possible to create a German syllable with /pʀ/ at the beginning, like ‘plump’. A syllable beginning with /pt/, however, is disallowed in German. Although the set-up of table 4.2 is similar to that of 4.1, you will notice that the allowable sequences are different. Thus, Table 4.1 Consonant clusters at the beginning of German syllables (adapted from Hall 2000:231). ʀ

p b t d k g pf ts f v s ʃ

l

+ + + + + + (+) − + (+) − +

+ + − (+) + + + − + (+) (+) +

m

n

− − − − − (+) − − − − (+) +

v

− − − − + + − − − − − +

p

− − − − + − − + − − − +

t

− − − − − − − − − − − +

k

− − − − − − − − − − − +

− − − − − − − − − − + −

Table 4.2 Consonant clusters at the end of German syllables (adapted from Hall 2000:235). ʀ l ŋ n m ç ʃ s f k t p

ʀ − − − − − − − − − − − −

l

+ − − − − − − − − − − −

n + + − − − − − − − − − −

m + + − − − − − − − − − −

ç

+ + − + − − − − − − − −

ʃ

+ + (+) + (+) − − − − − − (+)

s

+ + + + + + + − + + + +

f

+ + − + + − − − − − − −

k + + + − − − − + − − − −

t

+ + + + + + + + + + − +

p + + − − + − − − − − − −

Syllables and Word Stress  67 although /pt/ is not allowed at the beginning of German syllables, it is allowed at the end, for example, in a word like ‘head’. It is not necessary for you to memorize the details of tables 4.1 and 4.2. Many of the disallowed sequences in these tables—for example, the fact that /lk/ cannot occur at the beginning of a syllable but that it can occur at the end of a syllable in a word like ‘people’—can be accounted for by the sonority principle. 4.4. Building Syllables A typical German syllable has a vowel at its core. This leads us to our first rule for creating syllables in German: (6)

Every German syllable must contain a vowel.

The rule in (6) holds for the phonemic level of representation. At the phonetic level, sonorant consonants like nasals and /l/ can also function as the core of a syllable in German (see chapter 6 for more details). Rather than complicate our discussion to include these cases, we will keep the (simplified) rule in (6) and limit our discussion in this chapter to syllables that have a vowel as their core. Syllables in German can also have consonants on either side of the vowel. In (7), which is based on Wiese (2000:38), we show the basic syllable template for German, where the Greek letter sigma, σ, stands for syllable, C stands for consonant, and V stands for vowel. (7)



German syllable template σ C CVC C

According to the template in (7), a German syllable is created in such a way that it contains maximally five “slots” to fill, and these are a combination of vowels and consonants. Taken together with the phonotactic constraints in tables 4.1 and 4.2, this basic template allows us to determine that the syllables in (8) are possible, even if they do not actually occur in the German language. (8)

/glɔŋk/, /dʀʊlt/, /ʃmɪçs/, /kvapʃ/

By the same token, the segment strings in (9) are not potential syllables of German. (9)

*/bmɛkn/, */fvʏpl/, */gʀlts/, */vfœtm/

Although the simple syllable template in (7) can be helpful in making basic judgments about monosyllabic words, for the purposes of determining some important aspects of German syllabification, especially regarding the pronunciation of longer words, a more detailed syllable template like that in (10), based on Wiese 2000:44, will prove to be more useful.

68  Syllables and Word Stress (10) Detailed German syllable template σ

Onset

Rhyme

Nucleus



C

CV

Coda

X

C

In this syllable template, we can see that although the same basic elements are present as in (7), the hierarchical structure of the syllable has been fleshed out to include a number of additional constituents. The vowel and any consonants that follow it within the same syllable are referred to as the rhyme. Any consonants that come before the rhyme in the same syllable are placed in the constituent called the onset. This more detailed skeleton can help to explain why, for example, ‘cold’ rhymes with ‘old’ and not with ‘found.’ The rhyme is a unit that plays a role in language games, mnemonics, and poetry. It contains the nucleus (the vowel) and the coda (any consonants that appear after the vowel). You may wonder why the second segment in the nucleus in (10) is marked as neither a consonant nor a vowel but rather as X. This is because this slot in the syllable skeleton is more flexible than others. It can be a vowel, in which case it appears in the nucleus of the syllable. For example, it can be the second vowel of a diphthong, as in (11). (11)

Syllable structure of ‘brown’

Onset

Rhyme

Nucleus



/b

ʀ a

ʊ

Coda

n/

If a vowel is long, as in (12), it will fill both the V and X slots.

Syllables and Word Stress  69 (12) Syllable structure of ‘(garden) bed’

Onset



Rhyme

Nucleus

Coda

e

t/

/b

We show that a vowel takes up two slots by drawing two lines from it to the nucleus. Because long vowels fill two slots in the syllable skeleton, there is no need to also use the diacritic [ː] to mark length. The segment marked X in (10) can also be a consonant. When this happens, it appears in the coda, because the nucleus only contains vowels. In the word ‘child’, for example, /n/ is the segment that fills the X slot and therefore appears in the coda. (13) Syllable structure of ‘child’

Onset



/k

Rhyme

Nucleus

Coda

I

n d/

As the syllable structure of shows, a coda may contain two consonants when the vowel is short. We noted above that a syllable has maximally five slots to fill. Up to two of these form the onset, and three may be in the rhyme. There are a few exceptions to this generalization: word-initial syllables with onsets that take up more than two slots and word-final syllables with rhymes that take up more than three slots. These syllables all have one thing in common: they contain one or more of the segments /s/, /ʃ/, /t/, and /d/. We refer to these segments as being extrasyllabic, since they

70  Syllables and Word Stress do not fit inside the syllable template. Words like ‘slave’ and ‘beach’ are examples of syllables with onsets that take up more than two slots. (14) Syllable structure of ‘beach’ σ

Onset

Rhyme

Nucleus



/∫

t

ʀ

Coda

a

n

d/

Words like ‘mist’, ‘moon’, and ‘enemy’ are examples of syllables with rhymes that take up more than three slots. has a rhyme made up of a short vowel followed by three consonants. (15) Syllable structure of ‘mist’ σ

Onset

Rhyme

Nucleus



/d

ʊ

Coda

n

s

t/

contains a long vowel followed by two consonants, and has a diphthong followed by two consonants. Remember that long vowels and diphthongs take up two slots in the syllable template.

Syllables and Word Stress  71 (16) Syllable structure of ‘enemy’ σ

Onset

Rhyme

Nucleus



/f

a

Coda

n

I

d/

For more information on extrasyllabicity, see Wiese (2000:47–49). German syllables do not have to have a coda; they can end in a vowel. We refer to syllables that end in consonants as closed syllables and those that end in a vowel as open syllables. When we create the syllable structures for open syllables, we simply leave off the coda, as in (17). (17) Syllable structure of ‘May’ σ

Onset

Rhyme

Nucleus



/m

a

I/

The second rule for creating syllables in German deals with onsets: (18) Onset Maximization Principle (OMP) for German a. German syllables must begin with an onset. b. Onsets must be filled with the maximum number of allowable consonants before a vowel.

The first part of the OMP is important because it provides a place for the insertion of the glottal stop. You will recall that we did not include the glottal stop among the phonemes of German in chapter 3 because its distribution is predictable: it occurs word-initially before a vowel; within a compound before a word that begins with a vowel; and after a prefix if the following sound is a vowel. For example, it

72  Syllables and Word Stress does not occur in the phonemic representation of a word like ‘old’, /alt/, but it does in the phonetic representation, [ʔalt], and the first part of the OMP ensures that there is a slot in the syllable into which it can be inserted. You can see this in the syllable structure for the phonetic representation of in (19). (19) Syllable structure of ‘old’ σ

Onset

Rhyme

Nucleus



[ʔ

Coda

a

l

t]

Although the first part of the OMP makes onsets obligatory, they are not always filled. In a word like ‘to see’ /zeːən/, for example, the second syllable is not an environment for the insertion of a glottal stop, so the onset remains empty, as we see in (20). (20) Syllable structure of ‘to see’ σ

Rhyme

Nucleus



[z

e

Onset

Rhyme

Nucleus

e

Onset

σ

Coda

n]

In German, the glottal stop is never inserted into an empty onset in a syllable with schwa in its nucleus. The second part of the OMP helps determine which consonants belong to the coda of one syllable and which belong to the onset of a following syllable. In the word ‘liar’, for example, the second part of the OMP tells us that /gn/ must form the onset of the second syllable, since this is an allowable sequence of

Syllables and Word Stress  73 consonants (you can confirm this by consulting table 4.1): /lyː.gnəʀ/ (remember that a period is used in transcriptions to indicate syllable boundaries). In a word like ‘gardener’, on the other hand, the onset of the second syllable can only contain the consonant /n/, because neither /ʀtn/ nor /tn/ are acceptable onsets in German: */gɛ.ʀtnəʀ/, */gɛʀ.tnəʀ/, /gɛʀt.nəʀ/. As we noted at the beginning of this chapter, syllable structure is essential for determining where final obstruent devoicing occurs. Only obstruents that appear at the ends of syllables are devoiced. Consider related words like ‘cleverness’ /kluːghaɪ ̯t/ and ‘clever’ /kluːgə/ (the adjective with the ending ) in (21) and (22). (21) Syllable structure of ‘cleverness’ σ

σ

Onset

Rhyme

Nucleus



/k

l

Onset

Coda

u

Rhyme

Nucleus

g

h

a

Coda

t/

I

The /g/ serves as the coda to the first syllable of because /gh/ is not a possible syllable onset. Because /g/ is in syllable-final position, it will be devoiced: /kluːg.haɪ ̯t/ will be pronounced [kluːk.haɪ ̯t]. In , though, the OMP stipulates that /g/ must be the onset of the second syllable. Because it occurs syllable-initially, it will not be devoiced: /kluː.gə/ will be pronounced [kluː.gə]. (22) Syllable structure of ‘clever’ σ

Rhyme

Onset

Nucleus



/k

l

u

Rhyme

Nucleus

g

/

e

Onset

σ

74  Syllables and Word Stress One final topic that deserves discussion is the relationship between syllable boundaries and word boundaries. There are times when the syllabification of compound words does not conform to the OMP. For example, we might expect to syllabify a word like ‘basic right’ as */gʀʊn.dʀɛçt/. However, there is one rule that trumps even the OMP, and that is that syllable boundaries and word boundaries have to coincide (Benware 1986:85; Wiese 2000:65–67). (23) Syllable boundaries must coincide with word boundaries.

This means that the word has to be syllabified as /gʀʊnd.ʀɛçt/. (24) Syllable structure of ‘basic right’ σ

Onset

σ

Rhyme

Onset

Nucleus Coda



/g ʀ

ʊ

n

d

Rhyme

Nucleus

ʀ

ε

Coda

ç

t/

The /d/ in undergoes final devoicing, which is further confirmation that it occurs in the coda of the first syllable in the compound: [ gʁʊnt.ʁɛçt].

Steps to follow when syllabifying German words 1. Place each vowel (monophthong or diphthong) into a nucleus. 2. Within each word, build an onset to the left of each nucleus and fill it with the maximum number of allowable consonants. 3. Put all remaining consonants to the right of each nucleus in a coda. When you carry out the first step, use the syllable structure trees in (11) and (12) as a guide for putting diphthongs and long vowels into a nucleus. You will need to pay attention to the structure of words when you carry out the second step. If a word is a compound, make sure not to let a consonant at the end of one of the constituent words belong to the onset of a syllable in the following word. When you carry out the final step, if there are no consonants to the right of a nucleus, just leave off the coda for that syllable.

Syllables and Word Stress  75 4.5. Word Stress If you have ever counted out the syllables of a word, you have probably noticed that there is one syllable that is emphasized more than the others. We refer to this syllable as the one receiving primary word stress, which is marked in the following words with the symbol . This symbol is placed before the syllable with primary stress. Words like those in (25) all have the same stress pattern, a stressed syllable followed by an unstressed syllable, which we call trochaic stress. (25) ‘hammer’, ‘to sit’, ‘brother’, ‘girl’

Words like those in (26) have a stress pattern referred to as iambic, unstressed-stressed. (26) ‘lettuce’, ‘along’, ‘ban’, ‘talent’

You may have heard about iambic and trochaic stress if you have studied poetry. Word stress also plays an important role in everyday speech. It can distinguish the meanings of words. For example, the German words ‘to run over’ and ‘to circumnavigate’ differ phonetically only in the placement of word stress. If you look at the words in (27), all of which have three syllables and each of which contains stress on a different syllable, it appears as if stress assignment in German is random. (27) a. ‘parade’ b. ‘bakery’ c. ‘railroad’

In spite of what may appear to be unsystematic stress assignment in the examples in (27), stress assignment in German is largely systematic and predictable: exhibits default stress, contains the suffix , which is always stressed, and exhibits a stress pattern that is typical of compounds. In the sections that follow, we will provide you with a series of guidelines to determine where stress is assigned in German words. As you will see, you will need to be aware of the structure of words in order to assign stress. This is because the generalizations about stress assignment are sensitive to the structure of words. There are three types of words that are relevant for stress assignment: simplex words (these do not have prefixes or suffixes and are not compounds); complex words (words formed with prefixes and/or suffixes); and compound words (words created by combining two or more words). The focus of our discussion in the following sections is the placement of primary stress. However, when we discuss stress in compounds, we will introduce

76  Syllables and Word Stress the notion of secondary stress. As you will see, we need to recognize a secondary level of stress in order to account for long vowels that occur in compounds in syllables that do not bear primary stress. 4.5.1. Simplex Words Simplex words in German are typically stressed on the penultimate (second-tolast) syllable. (28) ‘rubber’, ‘account’, ‘canister’, ‘hibiscus’, ‘rhododendron’

This stress pattern is referred to as the default stress pattern, since this is the stress pattern that German speakers tend to give foreign words and foreign names when they are not aware of the stress pattern of the word in the source language. The stress assignment in the following words reflects this tendency. (29) ‘kamikaze’, ‘pick-up sticks (a game)’, ‘origami’, ‘safari’

We will mention one other important generalization that is useful for assigning stress in simplex words. Notice that the two simplex words in (26), and , do not exhibit default stress. They are stressed on the last syllable, not the second-to-last. Both words are loanwords, but we can predict their stress pattern without this information. They follow another rule of stress assignment that we will call the “super heavy syllable rule”: If a word-final syllable ends in a diphthong, a long vowel and at least one consonant, or a short vowel and at least two consonants, then it will receive primary stress. When we say “consonant,” we mean consonant phoneme. This rule accounts for the stress assignment in the following words—as well as in and . (30) a. ‘abbey’, ‘trouble’ b. ‘bandit’, ‘pirate’ c. ‘alarm’, ‘concert’

When you assign stress in simplex words, check first to see whether the last syllable of the word meets one of the criteria for the super heavy syllable rule. If it does not, then you should apply the default stress rule. 4.5.2. Complex Words We will describe stress in complex words by first looking at the way in which suffixes affect stress. Then we will talk about the role that prefixes play in determining word stress.

Syllables and Word Stress  77 4.5.2.1. Suffixes We can identify two very general types of suffixes: inflectional and derivational. Inflectional suffixes are those like adjective endings, verb endings, plural endings on nouns, and so on. These types of endings are used to create different forms of a word. Derivational suffixes are used to create new words. The derivational suffix , for example, creates nouns from verbs: ‘apartment’ is derived from ‘to dwell’; ‘rescue’ is derived from ‘to save’. Inflectional suffixes that contain a vowel and are capable of forming a syllable are always unstressed, since this vowel is /ə/, and /ə/ is never stressed in German. Inflectional suffixes typically do not change the stress pattern of the word to which they are attached. (31) ‘king’ ‘meager’

‘kings’ ‘meager’

Derivational suffixes fall into two basic classes with respect to word stress: those that do not bear stress and those that do. The majority of derivational suffixes that do not bear stress are Germanic in origin and do not affect the stress pattern of the word to which they are attached. (32) ‘dark’ ‘knowledge’ ‘summer’

‘darkness’ ‘science’ ‘summery’

Table C.1 in appendix C provides a list of common suffixes that do not bear primary stress. Several of these suffixes retain long vowels even though they do not bear primary stress (, , , and ). The word ‘valuable’, for example, with the suffix , is pronounced [ˈkɔstbaːɐ̯]. Phonetic transcriptions for these suffixes are provided in table C.1 so that you will know how to transcribe them phonetically. A small number of suffixes that do not bear stress can sometimes bring about a change in stress. One of these is . When this suffix is added to a complex word or a compound, it typically does not have an effect on stress. (33) ‘cheat’ ‘know-it-all’

‘deceitful’ ‘know-it-all’

However, when it is added to a multisyllabic simplex word, it can bring about a change in the stress pattern: (34) ‘chaos’ ‘Europe’ ‘Japan’

‘chaotic’ ‘European’ ‘Japanese’

78  Syllables and Word Stress Notice that this change in stress results in adjectives with a penultimate stress pattern, that is, with the default stress pattern found in simplex words. The derivational suffixes that do bear stress are all of Latin or French origin. These suffixes are stressed on the last syllable that does not contain [ə]. (35)

‘to graduate from’ ‘disgrace’ ‘Germanist’ ‘solidarity’

If a word contains several of these suffixes, only the last one is stressed. (36) ‘emotion’, ‘emotional’, ‘emotionality’

Table C.2 in appendix C provides a list of common suffixes that bear stress. 4.5.2.2. Prefixes Stress assignment in words with prefixes is sensitive to the part of speech of a word. Prefixes that are attached to verbs, for example, behave differently with respect to stress than those attached to nouns. We can identify three categories of verbal prefixes: 1) always stressed; 2) always unstressed; 3) stressed or unstressed. For the purposes of this discussion, we will refer to forms like in ‘to get in’ as well as those like in ‘to climb’ as prefixes. Strictly speaking, those like are not prefixes (they are not always bound to the rest of the verb), but we will call them prefixes, since this will make our discussion of stress less complicated. They are commonly referred to as separable prefixes. The list of verbal prefixes that are always unstressed (and never separated from the verbal stem) is short: [bə], [ʔɛmp], [ʔɛnt], [ʔɛɐ̯], [ gə], [fɛɐ̯], and [tsɛɐ̯]. These prefixes do not have any effect on the stress of the words to which they are attached. (37) ‘to ask’ ‘to need’ ‘to break’

‘to question’ ‘to use’ ‘to break into little pieces’

The prefixes , , , , , and can either be unstressed or unstressed. (38) ‘to supply with blood’ ‘to translate’ ‘to entertain’

‘to bleed through’ ‘to take across’ ‘to hold underneath’

As the examples here demonstrate, this difference in stress corresponds to a difference in meaning. There is also a correlation between stress and separability.

Syllables and Word Stress  79 When these prefixes are unstressed, they are inseparable; they are never separated from the verb stem. When they are stressed, they are separable. (39) a. b.

Sie übersetzte Werke aus dem Englischen. ‘She translated works from English.’ Er setzte uns ans andere Ufer über. ‘He took us across to the other bank.’

The verbal prefix , as well as with the meaning ‘too much’ and with the meaning ‘too little’, are unstressed unless followed by another unstressed prefix, in which case they are stressed. (40) ‘to disregard’ ‘to overtax’ ‘to underestimate’

‘to misunderstand’ ‘to overstress’ ‘to underpay’

All verbal prefixes other than those we have discussed so far are always stressed. That is, verbal prefixes other than those like and , for example (which are always unstressed), and and (which can be stressed or unstressed), are always stressed. Here are some examples of prefixes that fit into this category: (41)

‘to break off’ ‘to get out’ ‘to portray’ ‘to join in’ ‘to imitate’

Prefixes that are added to nouns and adjectives are typically stressed. The nominal and adjectival prefixes that are generally stressed are [ʔɛɐ̯ts], [mɪs], [ʔʊn], and [ʔuːɐ̯]. (42) ‘patriarch’ ‘failure’ ‘shallow’ ‘jungle’

‘ultraconservative’ ‘unpopular’ ‘unfaithful’ ‘ancient’

In order to determine the stress pattern of a word with a prefix, you will need to pay attention to its part of speech. For dealing with verbs, you should learn the short list of verbal prefixes that are always unstressed. You can find these in table C.3. It will also be useful to know which prefixes can be stressed or unstressed. If you encounter a verb with one of these, you will need to find it in a dictionary to determine whether it is stressed or unstressed—unless, of course, you already know the verb and its stress pattern. Table C.4 in appendix C provides

80  Syllables and Word Stress a list of common prefixes that are capable of bearing stress. Those in parentheses are used both with and without stress. The remaining prefixes in the table are always stressed. 4.5.3. Compound Words Compound words are created by combining two or more words. Although compounds can potentially be quite long ( ‘Australian long distance direct flight stopover specialists’, Donalies 2002:62), we will limit our discussion to two- and three-member compounds. There are two main types of compounds, copulative compounds and subordinate compounds. In copulative compounds, all members are of equal status (they could be conjoined with ‘and’), and each member is stressed. (43) ‘Schleswig-Holstein’ ‘black-red-and-gold’

In subordinate compounds, one member of the compound further specifies or defines another. In the subordinate compound ‘noodle soup’, for example, the first member, ‘noodle’, specifies the kind of ‘soup’. The typical stress pattern for subordinate compounds is stress on the first member. This is true of two- as well as three-member compounds. The internal structure of three-member compounds does not have a bearing on stress. For compounds with A(BC) structure, as well as for compounds with (AB)C structure, stress is on the A constituent. (44) ‘noodle soup’ ‘zoo veterinarian’ A(BC) ‘expressway network’ (AB)C

There is a special set of three-member compounds in which the B member is stressed: (45) ‘Red Cross nurse’ ‘touch-typing method’

These compounds differ in several ways from those with initial stress (Benware 1986:108): the A constituent is an adjective or a number and the (AB) constituent does not occur independently as a compound (*, *). When you do the exercises below that involve stress in compounds, you will need to keep in mind the relationships among the constituents in the compound. You will need to determine whether the constituents are equal in status, or whether one modifies another. If the compound has the structure (AB)C, you will also need to know whether the (AB) portion exists on its own as a compound. The answers to these questions will help you assign stress.

Syllables and Word Stress  81 You may have wondered what happens (if anything) to the stress pattern of a compound when it is shortened by turning it into an acronym. Although more than just compounds are turned into acronyms (many, like IDS, Institut für Deutsche Sprache, stand for phrases), you do not have to know anything about the source of an acronym in order to know how to stress it. Regardless of its source, it is always the last letter of an acronym that bears primary stress. (46) a. for ‘apartment-sharing community’ b.  for ‘German Democratic Republic’ c.  for ‘General German Automobile Club’

A final issue that we will look at in our discussion of compounds involves the notion of secondary stress. If you consider the pronunciation of the compounds in (47), you will see that long vowels appear in syllables that do not bear primary stress. (47) a. ‘streetcar’ /ˈʃtʀaːsənbaːn/ [ˈʃtʁaːsənbaːn] b. ‘elementary school’ /ˈgʀʊndʃuːlə/ [ˈgʁʊntʃuːlə] c. ‘rental apartment’ /ˈmiːtvoːnʊŋ/ [ˈmiːtvoːnʊŋ]

The in in (47a), for example, is realized as a long vowel, [aː], even though the syllable in which it occurs does not bear primary stress. Because long vowel phonemes that occur in unstressed syllables are realized as short vowels, we have to assume that the syllable bears some degree of stress. We identify this level of stress as secondary, and indicate it using a low vertical stroke, [ˌ]: [ˈʃtʁaːsənˌbaːn]. We will not require you to indicate secondary stress in compounds. However, if a vowel is long when it bears primary stress in a word used on its own, you will need to remember to transcribe it as long when it occurs in a compound, even if it does not occur in a syllable that bears primary stress. For example, the in ‘zone’ is long when the word occurs on its own, [ˈtsoːnə], so it is long in the compound ‘firing range’, even though it is not the first element in the compound and therefore does not bear primary stress: [ˈfɔɪ ̯ɐtsoːnə]. 4.6 . Advanced Topics 4.6 .1. Ambisyllabic Consonants A number of linguists take the position that a single intervocalic consonant in German that follows a short vowel is ambisyllabic, that is, that it belongs to the coda of the preceding syllable and to the onset of the following syllable (Vennemann

82  Syllables and Word Stress 1982, Giegerich 1985, Benware 1986, Wiese 2000). The /k/ in a word like ‘delicious’, for example, is assumed to be ambisyllabic, as illustrated in (48). σ

Onset

Rhyme

Nucleus

[l

σ

ε

Onset

Rhyme

Nucleus

Coda

k

]

a

(48)

One argument that supports this analysis is based on two facts about German phonotactics. First, word-final syllables in German do not end in a phonemically short vowel (with the exception of schwa); that is, /ɪ ʏ ɛ œ a ʊ ɔ/ do not occur word-finally in German. And second, word-initial syllables do not begin with /ə/. If we assume that these constraints on syllable structure hold not just at the edges of words but also word-internally, then /k/ in has to be ambisyllabic. It has to be in the coda of the first syllable (so that the syllable does not end in a short vowel) and in the onset of the second syllable (so that it does not begin with schwa). Ambisyllabicity can be indicated in phonemic and phonetic transcriptions by placing the symbol for a syllable boundary, a period, above the ̇ segment that is ambisyllabic: /lɛkəʀ/. As Hall (1992:52) points out, a flaw in this argument is the assumption that generalizations that hold for word edges also hold for syllable edges. He cites a number of studies that show that generalizations holding at word edges do not necessarily hold at syllable edges. In addition, there does not appear to be clearcut phonetic evidence for ambisyllabicity in German. Fischer-Jørgensen (1969) provides experimental evidence that intervocalic consonants after short vowels are generally somewhat longer than those after long vowels. Spiekermann (2000:66), on the other hand, finds the opposite tendency. We take the position that ambisyllabic consonants are not necessary for the treatment of syllable structure in German. 4.6 .2. Syllable Boundaries and Word Boundaries Before we provide some more information about the relationship between syllable boundaries and word boundaries, we will need to define the term “mor-

Syllables and Word Stress  83 pheme.” Put simply, a morpheme is the smallest linguistic unit that has meaning. The word ‘child’, for example, is made up of one morpheme. The plural form, ‘children’, is made up of two: and the plural ending . The verb ‘(you) destroy’ is made up of three: the prefix , the root of the verb, , and the inflectional suffix . We use the plus symbol to indicate morpheme boundaries: , . If a morpheme boundary is also a word boundary, as in a compound word like , we signal this by using the pound sign: . If you consider the syllable structure of a complex word like ‘organization’ (which is formed with the prefix ), you will see that the morpheme boundary after the prefix behaves like the word boundary in the compound ‘health resort’. (49) a. ‘organization’ /fɛʀ.aɪ ̯n/ [fɛɐ̯.ʔaɪ ̯n] b. ‘health resort’ /kuːʀ.ɔʀt/ [kuːɐ̯.ʔɔɐ̯t]

The /ʀ/ in , like the /ʀ/ in , cannot serve as the onset to the following syllable, which suggests that the morpheme boundary after a prefix blocks the OMP from applying. How do morpheme boundaries before suffixes behave with respect to the OMP? It turns out that morpheme boundaries before suffixes that begin with a consonant behave differently from those before suffixes that begin with a vowel. The boundary before the suffix , for example, prevents the OMP from applying, but not the boundary before the suffix . (50) a. ‘colorless’ /faʀb.loːs/ [faɐ̯p.loːs] b. ‘colored’ /faʀ.bɪg/ [faɐ̯.bɪç]

The /b/ in does not serve as the onset of the following syllable in , but it does in . The /b/ in undergoes final devoicing, which demonstrates that it occurs in a coda; the /b/ in does not, so it must be in an onset. In order to account for the fact that the boundaries after prefixes and those before consonant-initial suffixes behave like the boundaries between words in compounds when it comes to syllabification, Wiese (2000:66) argues that all of these boundaries are phonological word boundaries (indicated here with the symbol “ω”). Phonological words are phonological units within which certain phonological processes apply. In German, prefixes, roots, and suffixes that begin with a consonant followed by a vowel all function as phonological words. A suffix that does not begin with consonant followed by a vowel (, for example) is integrated into the preceding phonological word, as (51d) demonstrates (Wiese 2000:67).

84  Syllables and Word Stress (51)

a. (Ver)ω(ein)ω /fɛʀ.aɪ ̯n/ b. (farb)ω(los)ω /faʀb.loːs/ c. (Kur)ω(ort)ω /kuːʀ.ɔʀt/ d. (farbig)ω /faʀ.bɪg/

If we understand “word boundaries” in the rule in (23) to mean “phonological word boundaries,” we can account for the syllabification in complex words (words with prefixes and/or suffixes) as well as in compounds. In all cases, syllable boundaries coincide with phonological word boundaries. (For more information on phonological words, see Hall and Kleinhenz 1999 and Wiese 2000.) 4.6.3. Stress in Compounds When we talked about stress in three-member compounds above, we pointed out a special set of (AB)C compounds that have primary stress on the B constituent instead of the A constituent, which is the norm. A compound that is included in this set is ‘hot-water pipe’. You will recall that the A constituent in this type of compound is an adjective or a number and the (AB) constituent does not occur independently as a compound (*). There are two additional sets of compounds with primary stress on the B constituent. These compounds differ from those in the first set in that they have the structure A(BC). The first set of A(BC) compounds with stress on the B constituent designate specific days in the periods of time that involve Easter (), Pentecost ­(), and carnival (). (52) a. ‘Easter Sunday’, ‘Good Friday’ b. ‘Pentecost’, ‘Pentecost Monday’ c. ‘Shrove Tuesday’, ‘Ash Wednesday’

There are many other compounds that begin with words like , , and but do not designate specific days and therefore have primary stress on the A constituent: (53) a. ‘Easter holiday’, ‘Easter egg’ b. ‘Pentecost holidays’, ‘peony’ c. ‘carnival period’, ‘carnival costume’

The second set of A(BC) compounds with stress on the B constituent are those that begin with , , and . (54) a.  ‘Federal Labor Court’, ‘Federal Intelligence Service’ b.  ‘state unemployment office’, ‘state constitutional court’

Syllables and Word Stress  85

c.  ‘International Women’s Day’, ‘Food and Agriculture Organization of the United Nations’

It is important to be aware of the structure of compounds that begin with , , and , since those with the structure (AB)C will have primary stress on the A constituent, not the B constituent. (55) a. ‘President of the Bundestag’ b. ‘law dealing with regional media’ c. ‘space research’

Although primary stress in subordinate compounds is typically on the first constituent of the compound, you will no doubt come across three-member compounds with primary stress on the second constituent like those we have presented here. When you do encounter such compounds, their stress pattern should come as no surprise. Exercises 4.1. How many syllables are in the following German words? a. b. c. d. e. f. g. h. i. j. k. 4.2. Compare the English and German words below. How many syllables are in each? a.

b.

c.

d.

e.

f.

g.

h.

86  Syllables and Word Stress 4.3 Circle the most sonorous segment in each of the following sets. a. /t k s m l/ b. /a l i f n/ c. /d ʀ m v l/ d. /p z ŋ ʊ/ e. /u s g a k/ f. /b n d g t/ 4.4. Which of the following is a possible word in German? Explain why those that you have determined are not possible are indeed not possible. a. b. c. d. e. f. g. h. i. j. 4.5. Explain why each of the following words is possible in German according to the sonority principle. a. b. c. d. e. f. 4.6. Refer to the set of possible onset clusters in German syllables in table 4.1. Come up with German words containing each of the possible clusters except for those marked with “(+).” 4.7. Refer to the set of possible coda clusters in German syllables in table 4.2. Come up with German words containing each of the possible clusters except for those marked with “(+).” 4.8. Create syllable structure trees for the phonetic transcription of the following monosyllabic German words. a. b. c.

Syllables and Word Stress  87 d. e. f. g. h. i. j. 4.9. Create syllable structure trees for the phonetic transcription of the following multisyllabic German words. a. b. c. d. e. f. g. h. i. j. 4.10. Transcribe the following words phonetically, using a period to mark syllable boundaries. You may wish to create syllable structure trees to assist you. a. b. c. d. e. f. g. h. i. j. k. l. 4.11. Stress in simplex words. Mark the stressed syllable in the following words and indicate which rule applies: super heavy syllable rule; default stress rule. a. b. c. d.

88  Syllables and Word Stress e. f. g. h. i. 4.12. Stress in words with suffixes. Mark the stressed syllable in the following words. a. b. c. d. e. f. g. h. i. j. k. l. m. n. 4.13. Stress in words with prefixes. Mark the stressed syllable in the following words. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. p. ‘to insinuate’ q. ‘to breathe deeply’ r. ‘to contradict’ s. ‘to embrace’ t. ‘to translate’

Syllables and Word Stress  89 4.14. Stress in simplex and complex words. Mark the stressed syllable in the following words and then transcribe each word phonetically. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. 4.15. Stress in two-member compounds. Mark the stressed syllable in the following words. If one of the members of a compound is a multisyllabic simplex or complex word, use the rules for simplex and complex words to determine which syllable of that word to stress. a. b. c. d. e. f. g. h. i. j. 4.16. Stress in three-member compounds. Determine whether the following compounds are A(BC) or (AB)C compounds. Then mark the stressed syllable in each compound. a. b. c. d. e. f. g.

90  Syllables and Word Stress h. i. j. 4.17. Word stress. Mark the syllables with primary stress in the following words. Make sure to pay attention to the structure of each word. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. p. 4.18. Transcribe the following words phonetically and mark the syllables that bear primary stress. Remember that if a vowel is long when it bears primary stress in a word used on its own, you will need to transcribe it as long when it occurs in a compound, even if it does not occur in a syllable that bears primary stress. a. b. c. d. e. f. g. h. 4.19. Advanced: Syllable boundaries and word boundaries. Transcribe the following words phonetically, using a period to mark syllable boundaries. a. b. c. d.

Syllables and Word Stress  91 e. f. g. h. i. j. 4.20. Advanced: Stress in compounds. Mark the syllables with primary stress in the following words. Make sure to pay attention to the structure and meaning of each word. a. b. c. d. e. f. g. h. i. j. k. l.

This page intentionally left blank

Chapter 5

Prosody

5.1. Introduction You may be familiar with the term “prosody” in the context of poetry, where it is used to refer to the rhythmic structure of verse. We also use this term in linguistics, where it refers not just to the rhythm of speech, but also to stress (relative emphasis given to certain syllables in a word or to the syllables of certain words in a phrase or sentence) and intonation (changes in pitch in spoken utterances). German language textbooks usually pay little attention to German prosody at the level of the utterance beyond perhaps basic discussions of the role of intonation in distinguishing declarative utterances from questions. This may be because many consider prosody to be an add-on. However, it is important for learners of German to know that prosody can be used in a number of ways in German that have a significant effect on meaning. We can divide the uses into two basic types: paralinguistic and linguistic. “Paralinguistic” is a term used to describe those functions of language that do not affect the core meaning of an utterance. When we use intonation to show emotions or attitudes, this is a paralinguistic use of intonation. For example, consider how a speaker of German might vary his intonation of (1) based on the emotions he may want to express. (1)



Die Kinder arbeiten am Computer. ‘The children are working on the computer.’

Generally, we expect a speaker to produce this utterance as a simple declarative. That is, we expect falling intonation with the main stress of the sentence on the final content word, ‘computer’. However, someone who walked into the room ready to use the computer for work might utter the same statement in surprise, which may involve rising intonation. Another speaker, excited that the children are making use of new software that she just purchased, might say the same utterance with “happy” prosody in her voice, which may mean that she  speaks with more intonational variation (with higher pitch than usual on , for example). There are times when prosodic cues can distinguish meaning at the utterance level. When this is the case, we say that the prosodic cues are linguistic. The intonation contour used to signal the difference between a declarative utterance and 93

94  Prosody a question makes a grammatical contrast. Consider the difference between ‘The sun is shining’ and ‘The sun is shining?’ The declarative utterance is spoken with falling intonation, and the question is spoken with rising intonation. When speakers want to contrast one piece of information with another, they rely on prosodic cues to highlight that piece of information. Consider how the meaning of (2a) differs from that of (2b). The stressed syllables in these utterances are capitalized. (2) a. b.

Ich habe den SCHWARzen Pulli gemeint (. . . nicht den BLAUen.) ‘I meant the black sweater (. . . not the blue one.)’ Ich habe den schwarzen PUlli gemeint (. . . nicht den schwarzen MANtel.) ‘I meant the black sweater (. . . not the black coat.)’

If you consider the differences in stress assignment in the utterances along with the potential continuations that are in parentheses, you will see that the sentences differ in their meaning. One additional example of the linguistic use of prosody comes from studies looking at ambiguity. Perhaps the clearest example of the use of prosodic cues to distinguish the meaning of potentially ambiguous utterances comes from the realm of mathematics. Consider the cues a speaker would use to distinguish the mathematical equation in (3a) from the equation in (3b). (3)

a. (3 × 5) − 2 b. 3 × (5 − 2)

Even though the numbers are the same in (3a) and (3b), the final answer in these examples differs according to the prosodic cues a speaker uses. A speaker uttering (3a) would likely use rising or level intonation when uttering ‘five’. In addition, he or she would probably produce a pause after to indicate that is a unit that belongs together. Alternatively, a speaker who utters (3b) would most likely use rising or level intonation on, and a pause following, ‘three times’. This would indicate to the listener that the first unit has been uttered. The final unit in both (3a) and (3b) would most likely be produced with falling intonation. To determine whether the use of prosodic cues is paralinguistic or linguistic, consider whether the use of various cues affects the core meaning of what is being uttered. If prosody changes the core meaning, then the use of prosodic cues is linguistic. 5.2. Rhythm Rhythm is a feature of speech that involves the perceived regularity with which syllables recur. In German, like English, rhythm comes from the patterning of

Prosody  95 stressed and unstressed syllables. Researchers group German and English together as stress-timed languages, which contrast with syllable-timed languages like French or Spanish. In stress-timed languages, stresses (stressed syllables) are the important units of rhythm; in syllable-timed languages, syllables play this role. It was once argued that stress-timed languages are isochronous, meaning that speakers of stress-timed languages take the same amount of time to produce any number of unstressed syllables that occur between one stressed syllable and the next. Thus, in an utterance like (4a), where each monosyllabic word is stressed, each word/syllable should take roughly the same amount of time to produce. The vertical lines indicate approximately equal intervals of time. (4)

a.  | drei b.  | einhundert c.  | Lügen und

| Glas | Flaschen | unverschämte Ver-

| Wein | Weißwein | leumdungen

| | |

In (4b) and (4c), it should take a speaker roughly the same amount of time between the stressed syllables, even though other syllables intervene. (See Fox [2002:87– 91] for a discussion of isochrony in English.) Even though the notion of isochrony does not hold up in terms of absolute duration of time between stressed syllables, there are two facts involving the duration of stressed syllables that are important to keep in mind. The first is that the duration of stressed syllables is longer than that of unstressed syllables. The second is that a stressed syllable is longer if it is followed by a stressed syllable than when it is followed by an unstressed syllable (Dasher and Bolinger 1982). It is important for second language learners, who often have difficulty reducing unstressed vowels, to work on activities that encourage them to reduce the length of vowels between stressed syllables. One topic of interest in an investigation of rhythm in German is the role that secondary stress plays in the pattern of stressed and unstressed syllables. Recall that in chapter 4 we identified a level of secondary stress to account for long vowels that occur in compounds in syllables that do not bear primary stress. The vowel in ‘baby carriage’, for example, is realized as long because it occurs in a syllable with secondary stress: [ˈkɪndɐˌvaːgən]. In a discussion of rhythm in German, Kaltenbacher (1998:26) argues that secondary stress placement is determined to a large extent by the principle of Alternation: secondary stresses are placed in German in a way that achieves the most regular alternation of stressed and unstressed syllables possible. She finds evidence for this in the formation of rhymes, which in German are constructed using the sequence of sounds beginning with the last stressed vowel in the line of verse. Syllables that are typically unstressed can receive secondary stress and participate in rhymes if they do not immediately follow a stressed syllable, as we see in the following lines of verse (cited in Kaltenbacher 1998:26):

96  Prosody (5)

a. Die Wespe mit vergnügtem Sinn // Betrachtet sich als Siegerin. (Wilhelm Busch) b. Der Architekt jedoch entfloh // nach Afri- od- Ameriko. (Christian Morgenstern)

In (5a), for example, the suffix in ‘victor’, although normally unstressed, can receive secondary stress and rhyme with in the word ‘sense’ because an unstressed syllable occurs between it and the first syllable of the word, which has primary stress: [ˈziːgəˌʁɪn]. 5.3. Phrasal Stress If you compare the compound words in the (a)-column in (6) and the corresponding noun phrases in the (b)-column, you will see that the placement of stress in phrases differs from that in compounds (the first two sets of examples are from Wiese 2000:303). (6) a. die ˈKrankenschwestern  ‘the nurses’ die ˈArmenhäuser ‘the poorhouses’ die ˈAltenheime ‘the old people’s homes’ die ˈGrünenabegeordneten ‘the Green  representatives’

b. die kranken ˈSchwestern ‘the  sick sisters’ die armen ˈHäuser ‘the poor houses’ die alten ˈHeime ‘the old homes’ die grünen ˈAbgeordneten ‘the  inexperienced representatives’

The compounds have primary stress on their initial element, but the phrases have primary stress on their final element. The first syllable of the adjective ‘inexperienced’ in the phrase is also stressed (the long vowel /yː/ is not shortened), but it only has secondary stress. The first syllable of is the most prominent syllable in the phrase. If you consider the phrases in (7) through (9), you will see that it is the last content word in a phrase (noun, verb, adjective, adverb) that receives the primary stress. We assume that these phrases are uttered under neutral circumstances. You can imagine the whole phrase as being the answer to a question. For example, the noun phrase in (7a) could be the answer to the question . (7)

Noun phrases a. der erste ˈTag ‘the first day’ b. der erste Tag des ˈMonats ‘the first day of the month’ c. der erste Name auf der ˈListe ‘the first name on the list’

(8)

Prepositional phrases a. durch die ˈLuft ‘through the air’ b. dem Arˈtikel nach ‘according to the article’ c. um der ˈLiebe willen ‘for the sake of love’

Prosody  97 (9)

Adjective phrases a. sehr ˈklug ‘very smart’ b. dem Leser beˈkannt ‘known to the reader’

The prepositional phrases in (8b) and (8c) show that it is not always the last word in a phrase that receives primary stress. In both cases, the last word in the phrase is a preposition, which is not a content word and therefore does not bear primary stress. Note that we use the term “preposition” as a cover term, since ‘according to’ in (8b) is in fact a postposition and ‘for the sake of ’ in (8c) a circumposition. The placement of primary stress in verb phrases differs somewhat from the placement of stress in the phrases above. (10)

Verb phrases a. einen neuen ˈWagen kaufen ‘to buy a new car’ b. ˈmüde sein ‘to be tired’ c. in die ˈBerge fahren ‘to drive to the mountains’

In each of the examples in (10), the verb is the final content word but it does not bear the primary stress in the phrase. An exercise at the end of this chapter will provide more information about stress placement in verb phrases and give you the opportunity to formulate a rule that accounts for the data. 5.4. Sentence Stress and Focus In each simple sentence there is one syllable that receives the greatest emphasis. We call this syllable the nucleus, and we say that it receives the nuclear stress. Refer to the exchange in (11), where each nucleus is highlighted via capitalization. (11)

A: Guten TAG! Herzlich Willkommen in unserem HoTEL. Wie kann ich Ihnen HELfen? ‘Hello! Welcome to our hotel. How can I help you?’ B: Guten TAG! Wir suchen ein ZImmer. ‘Hello! We’re looking for a room.’ A: Für wie viele PerSOnen? ‘For how many people?’ B: ZWEI Personen. Meine FRAU und ich. ‘Two people. My wife and I.’ A: Suchen Sie ein Zimmer MIT oder OHne Bad? ‘Are you looking for a room with or without a private bath?’ B: MIT Bad. ‘With a private bath.’ A: GUT. Wie ist Ihr NACHname? ‘Good. What is your last name?’ B: SCHMIDT. Bernhard SCHMIDT.

98  Prosody In the exchange in (11), a hotel clerk is speaking with a customer, Mr. Schmidt. In the first line, the clerk begins the conversation with no shared information. In the interactions that follow, each exchange involves some bit of new information. Each speaker highlights—or puts into focus—new information relative to the rest of the information in the utterance. For example, in the exchange about the private bath, both speakers place emphasis on prepositions. Although we might not expect prepositions to be highlighted in a neutral context, in a situation in which one preposition is contrasted with another, emphasizing the preposition relative to the other words in the utterance focuses attention on the most important information. 5.4.1. Broad Focus “Broad focus” is the term we use to characterize sentences in which no particular word is highlighted relative to any other. That is, all of the information in the utterance is new for the listener. Sentences spoken with broad focus can be uttered out of the blue or in response to a question like ‘What is happening?’ In this context, the speaker evaluates all of the elements as equally important. In all-new utterances, it is often the last content word that receives the nuclear stress. Content words include nouns, verbs (but not auxiliaries or modal verbs), adjectives, and adverbs. Function words, on the other hand, which include articles, pronouns, prepositions, and conjunctions, do not typically bear nuclear stress in all-new utterances. Consider the placement of the nucleus in the all-new utterances in (12), which are titles of children’s books by Alexander Steffensmeier. (12) a. b. c.

Liselotte macht URlaub. ‘Liselotte is on vacation.’ Liselotte versteckt sich. ‘Liselotte is hiding.’ Liselotte ist krank. ‘Liselotte is sick.’

It is not always the case that the last content word in all-new utterances bears the nuclear stress. Consider what happens when we put the examples in (12) into the perfect tense: (13) a. Liselotte hat URlaub gemacht. ‘Liselotte was on vacation.’ b. Liselotte hat sich versteckt. ‘Liselotte hid.’ c. Liselotte ist krank gewesen. ‘Liselotte was sick.’

Prosody  99 In (13a) and (13c), the last content word is the main verb in participle form ( ‘made’; ‘been’), but it is the preceding content word that bears the nuclear stress. How can our description of nuclear stress in utterances with broad focus be refined to account for the placement of stress in the utterances in (13) as well as those in (12)? We will not answer this question here. Instead, we will save it for the exercises at the end of this chapter in order to give you the opportunity to work with language data and discover generalizations about nuclear stress placement on your own. For more information about nuclear stress placement beyond what we present in the exercises, see Uhmann (1991). 5.4.2. Narrow Focus There are times when a speaker chooses to highlight one element of an utterance relative to the rest. In this case we say that the utterance has narrow focus. An utterance may have narrow focus because the highlighted information is new in the discourse context (the speaker is distinguishing new and given information). German speakers can distinguish new from given information in several ways. For example, discourse-new items are often identified through the use of indefinite articles, as in (14), and discourse-given information is often marked through the use of definite articles or pronouns, as in (15). (14) Samuel trägt einen RUCKsack. ‘Samuel wears a backpack.’ (15) A: Was ist mit dem Rucksack passiert? ‘What happened to the backpack?’ B: Samuel TRÄGT ihn. ‘Samuel is wearing it.’

The new information in example (14) is ‘a backpack’, and in example (15), because the discourse context is provided, we know that the new information is either or ‘is wearing’. If both are new, the speaker would be more likely to highlight the verb, since it is the content word closest to the end of the sentence. Because the backpack is referred to in the question asked in A, it is natural for the speaker in B to use the pronoun ‘it’ rather than the full noun to indicate that he or she is referring to the previously mentioned backpack. Given information—especially a pronoun—typically does not receive nuclear stress. Even when the given information is a noun, it does not receive nuclear stress. Consider the exchange in (16). (16) A: Spielt Samuel mit dem Rucksack? ‘Is Samuel playing with the backpack?’

100  Prosody B: Nein, er TRÄGT den Rucksack. ‘No, he’s wearing the backpack.’

In this example, the verb ‘is wearing’ is the new information, and everything else in the sentence is given, so the speaker emphasizes the verb. In many utterances with narrow focus, the information that is focused is new. In example (17), the focused information, ‘juice’, has not been mentioned in the question that precedes it and is therefore new information. (17) A: Was hast du getrunken? ‘What did you drink?’ B: Ich habe SAFT getrunken. ‘I drank juice.’

Not all focused information is new, however. There are times when a speaker places emphasis on an element that has already been mentioned. Consider example (18), in which the focused element is one of those mentioned in the question. (18) A: Hast du Saft oder Milch getrunken? ‘Did you drink juice or milk?’ B: Ich habe SAFT getrunken. ‘I drank juice.’

In German, there are two options for indicating narrow focus. The first is making use of prosodic cues, as in (17) and (18). The second is fronting, that is, moving the focused element to the first position of the sentence before the verb. Often the two happen together. In (18), for example, a speaker could also respond by saying ‘It’s juice that I drank.’ It is difficult—if not impossible—to distinguish acoustically between a response with direct object focus like (19) and one with broad focus like (20). Context (in this case, the questions provided in [19A] and [20A]) tells us whether the direct object is being focused, as in (19B), or whether the sentence contains broad focus, as in (20B). It is quite rare that a speaker would ever produce a sentence with narrow focus out of the blue. (19) A: Was schreibst du? ‘What are you writing?’ B: Ich schreibe einen BRIEF. ‘I am writing a letter.’ (20) A: Was passiert? ‘What’s happening?’ B: Ich schreibe einen BRIEF. ‘I am writing a letter.’

Prosody  101 5.5. Intonation Intonation is the melody, or tune, of an utterance. The intonation contour of an utterance is the pattern of pitch changes that a speaker produces. Acoustically, we can measure pitch via fundamental frequency (F0, the acoustic correlate of pitch). One of the most basic ways to think about intonation is to consider whether a sentence rises or falls at its end. When you learned German (or heard others talk about English), you might have heard that the intonation of statements is falling and that of questions is rising. This linguistic use of intonation is grammatical: it distinguishes between statements and questions. In broad terms, this may have worked for you. In reality, though, the picture is a bit more complex than this, as you will see in the following sections. The overview of German intonation that we present in this chapter is a basic guideline. It is important to note, though, that native speakers produce intonation contours with a great deal of variability. This can depend on the dialect they speak, the register they are using at a given time (formal, informal), and level of education, to name just a few potential influences. 5.5.1. Intonation Phrases An important role of intonation is phrasing, or dividing speech into chunks. The relevant unit that we speak about when we discuss intonation is the intonation phrase, which is also referred to as the tone group, tone unit, sense group, or breath group. An intonation phrase contains a nucleus, which we discussed above, along with the information that comes before it (the prenuclear information) and the words that come after it (sometimes referred to as the tail or the postnuclear domain). To illustrate this, consider the statements in (21), each of which contains one intonation phrase. The nucleus is capitalized in each of the examples. (21)

a. JA. b. KURT kommt. c. Er spricht mit KURT. d. Maria ist mit KURT gefahren.

Examples (21a) and (21b) do not contain any prenuclear information, and examples (21a) and (21c) have no information after the nucleus. Only example (21d) contains both prenuclear and postnuclear information. While research on intonation has traditionally focused on the nucleus and the information that follows, recent work has pointed to the importance of the prenuclear domain in the role it plays in conveying meaning. Petrone and Niebuhr (2014), for example, provide experimental data showing that speakers of Northern Standard German are able to distinguish between statements and questions on the basis of the intonation contour of the prenuclear portion of an intonation phrase. For more information on the prenuclear domain, see Féry (1993) and Gussenhoven (2004).

102  Prosody All of the examples in (21) are basic declarative sentences, and in a neutral context they would be produced with a falling contour. That is, the pitch in each sentence would drop after the nucleus was produced. We will discuss falling contours more below when we discuss the intonation of statements. Speakers often produce utterances that contain more than one intonation phrase. Even though there are a number of potential ways of determining the boundaries of intonation phrases, for our purposes, we will consider three important cues to mark an intonation phrase: a) the final syllable is lengthened; and it is followed by b) a pause and/or c) the resetting of the pitch contour. Consider the example of the mathematical equation in (3) above, where the boundaries between groups of numbers are indicated in writing via parentheses. A speaker may make use of any or all of these cues (lengthening the final syllable, pausing, resetting the pitch contour) to indicate boundaries like these. 5.5.2. Multiple Intonation Phrases Complex sentences, because they are longer and because of their grammatical complexity, usually contain more than one intonation phrase. To separate one phrase from another, speakers rely on prosodic cues. One very important use of prosodic cues is to indicate the intended syntactic structure of an utterance. Take an example like (22). (22) Lena verfolgt den Mann mit dem Motorrad. (Féry 1993:125) ‘Lena follows the man with the motorcycle.’

A sentence like this is syntactically ambiguous. Without context, it is impossible to determine who is traveling with the motorcycle. Depending on the intonational cues a speaker uses, it could be either Lena or the man. A speaker who wants to indicate that the man has the motorcycle would produce a prosodic break between the verb, ‘follows’, and the object, ‘the man’. This break, indicated through the use of , may involve a rise in pitch or lengthening of the verb and/or a pause before the direct object. The result is that the direct object and the following prepositional phrase, ‘with the motorcycle’, are produced as one chunk, as in (23a). If a speaker wants to indicate that Lena is driving the motorcycle, the phrase boundary would appear between ‘man’ and ‘with’, as in (23b). (23) a. Lena verfolgt // den Mann mit dem Motorrad. b. Lena verfolgt den Mann // mit dem Motorrad.

Less obvious uses of intonational cues to divide speech into phrases include the indication of items in a list, coordinate clauses, subordinate clauses (except those with ‘that’ and ‘whether’), and parenthetical expressions.

Prosody  103 (24) a. b. c. d.

Ich kaufe Bananen, Äpfel, Salat und Tomaten. ‘I’m buying bananas, apples, lettuce, and tomatoes.’ Die Studenten haben viel gelesen, aber sie haben nichts verstanden. ‘The students read a lot, but they understood nothing.’ Unser Nachbar arbeitet im Garten, wenn er Zeit hat. ‘Our neighbor works in the garden when he has time.’ Die Schule ist, wie jeder weiß, viel zu klein. ‘The school is, as everyone knows, much too small.’

Intonational cues also help distinguish restrictive (defining) relative clauses from nonrestrictive (nondefining) relative clauses. A restrictive relative clause, as the example in (25) shows, provides information that defines the noun it modifies; it provides information that is necessary for accurate identification of the noun. A restrictive relative clause in German does not form a separate intonation phrase; there is no separate intonation phrase before the relative pronoun in (25). (25) Der Wagen, den Michael fährt, sieht teuer aus. ‘The car that Michael’s driving looks expensive.’

A nonrestrictive relative clause, on the other hand, simply provides extra information about its antecedent, as the example in (26) demonstrates. Nonrestrictive relative clauses are like parenthetical expressions, and like parenthetical expressions, they form a separate intonation phrase. (26) Der BMW i8, der ein Hybridfahrzeug ist, wurde von BMW entwickelt. ‘The BMW i8, which is a hybrid vehicle, was developed by BMW.’

In many situations, punctuation can assist you in the production of intonation phrases. A comma often signals a separate intonation phrase—but not always. One instance in which there is no separate intonation phrase before the comma is in the case of subordinate clauses with or , as in (27). (27) Der Mann hat gesagt, dass er viel zu tun hat. ‘The man said that he has a lot to do.’

There is also no separate intonation phrase before the comma that sets off a restrictive relative clause. Notice that English differs from German in the punctuation of relative clauses. As the English translations of the relative clauses in (25) and (26) show, English uses commas to distinguish between restrictive and nonrestrictive relative clauses. In English restrictive relative clauses there is no comma and no separate intonation phrase; in nonrestrictive clauses there is a comma and a separate intonation phrase. One additional mismatch between commas and intonation phrases in German involves lists. Although the last item in a

104  Prosody list ­constitutes a separate intonation phrase, it is not set off by a comma in German, as the example in (24a) demonstrates. 5.5.3. Intonation Contours It is common for nonfinal intonation boundaries to remain level or to rise in German. The continuation rise, as it is called, commonly indicates that a speaker will continue his or her utterance. Sentence-final phrases like in (24b) and in (24c) are produced with a falling contour. One generalization that can be made about intonation contours is this: falling contours tend to signal finality, and rising contours signal that information is forthcoming or that information is being requested from the listener (as in the case of a yes-no question). We will now return to the basic principle that is so often disseminated in German language classes: intonation falls in statements and rises in questions. Although German speakers produce yes-no questions like ‘Do you come from Hamburg?’ with rising intonation, they usually produce a question like ‘Where do you come from?’ with falling intonation. Thus, the generalization regarding the contours of German questions has to be amended: yes-no questions are often produced with rising intonation, but questions that begin with question words are often produced with falling intonation. There are also instances when speakers produce German utterances with declarative syntax and rising intonation, as in the example ‘The sun is shining?’ provided above. In this instance, rising intonation is required if the speaker intends to ask a question. Otherwise, the word order would signal that the speaker was making a statement. One final type of utterance that contains falling intonation is the imperative. Consider the commands in (28).

Standard intonation contours in German Falling intonation is typically used in statements and questions that begin with question words. Statements: Wir fahren morgen in den Urlaub. ‘We’re going on vacation tomorrow.’ Questions that begin with question words: Wann fahren wir in den Urlaub? ‘When are we going on vacation?’ Rising intonation is typically used in yes/no questions. Yes/no questions: Fahren wir morgen in den Urlaub? ‘Are we going on vacation tomorrow?’

Prosody  105 (28)

a. b. c. d. e.

MACH das nicht. Komm MIT. Sei nicht so LAUT. Öffnen Sie bitte das FENster. Rufen Sie mich AN.

When commands are produced with falling intonation, there is a sense of finality. There are also instances when a speaker might produce an imperative with a rise. Fox (1984:64) notes that when this happens, it adds an element of “appeal” that “turn[s] the command into a request” or that “add[s] a note of urgency, encouragement, [or] impatience.” Similarly, Fox reports that questions containing a question word that are spoken with rising intonation also add an element of appeal. Such questions might sound more friendly or polite. In addition, speakers sometimes produce questions with rising intonation when repeating questions that are posed to them (Fox 1984:63). Consider example (29), which could be the response to a question posed with level or falling intonation: ‘How old is the president?’ (29) Wie alt ist der Präsident? Das ist eine gute Frage. ‘How old is the president? That is a good question.’

In this instance, the speaker in (29) could produce the question with rising intonation either as an indication that he or she is especially interested in the question but does not have the information or to ask for confirmation (with a meaning something like ‘You’re wondering how old the president is, right?’). Few utterances end with level intonation. Fox (1984: 68–69) indicates that short, often formulaic or ritualistic, utterances like greetings may be spoken with level intonation, and such utterances are generally interpreted to be “cheerful” and “unassertive.” If a speaker uses level intonation in other circumstances, he or she may come across as bored, or it may seem as though the information in the utterance is “so self-explanatory as to be hardly worth mentioning.” Some examples of ritualistic utterances spoken with level intonation include ‘Nice!’ and ‘That’s right!’ As you have seen, there are clear reasons for making use of one intonation contour or another. If you decide not to make use of the expected (default) contour, you should do so with a particular goal in mind. If not, a listener might be confused as to the meaning of your utterance. 5.6. Advanced Topics Over time, researchers and teachers have proposed a number of approaches for transcribing intonation contours. Phonological transcription systems tend to focus on patterns of rises and falls throughout utterances. Early systems used in the United States differed somewhat from those in Britain. The American system

106  Prosody advocated by linguists such as Trager and Smith (1951) was based on a meaningbearing unit called the toneme, a pitch phoneme. Four levels of tonemes were proposed, each of which had a number (1, 2, 3, 4), and utterances were classified according to the patterning of tonemes, as in (30), based on Moulton 1962:131. (30) 2 ˌes

˚3 ˈregnet ˈheute

1↓ ˈAbend

In this system, lower numbers (1 and 2) corresponded to lower pitch levels. The diacritic [˚] that appears before the 3 indicates that the verb in this utterance receives the nuclear stress. Because the toneme over the verb has the highest value, this is an indication that the pitch rises at that point. The numbers 2 and 1 at the beginning and end of the sentence indicate that the sentence begins relatively low in the speaker’s pitch range and ends at an even lower level. The downward facing arrow at the end provides an indication that the pitch should fall at the end of the utterance. Intonation contours transcribed according to the British School model show pitch modulation via a series of configurations. Each syllable is annotated with a dot, and movement within the syllable receiving the nuclear stress (the nucleus) is shown using a dynamic pitch contour (a line). Stressed syllables are annotated with larger dots. An example is shown in figure 5.1. Today one of the most widely used systems for phonological transcription of German intonation contours is GToBI, which stands for German Tones and Breaks Indices (Grice and Baumann 2002), a system that is based on the English ToBI model (Pierrehumbert 1980; Pierrehumbert and Beckman 1988). These models, which are referred to as autosegmental-metrical models (Ladd 1996; Pierrehumbert 1980), assign prominence within the unit of the intonation phrase. The overall contour, or tune, of the utterance exists separately from the utterance itself. At certain points—for example, in the case of stressed syllables (pitch accents) and at the right edge of the phrase (boundary tones)—the tune is anchored to the text. Pitch accents (*) and boundary tones (%) are marked according to their relative pitch height (H for high and L for low). An example of a GToBI transcription is provided in (31), based on Grice and Baumann 2007:45. Stressed syllables appear with a starred tone (for example, L* or H*), and boundary tones at the ends of Prehead

Head

Nucleus

Mag- da-

LE- na hat ein

HAUS

Tail

ge-

kauft

Figure 5.1 British School transcription of ‘Magdalena bought a house’ (adapted from Grice and Baumann 2007:43).

Prosody  107 sentences are transcribed as H% or L%. Nuclear stress falls on the last pitch accent, in this case . (31) MagdaLEna hat ein HAUS gekauft. | | | L* H* L-%

‘Magdalena bought a house.’

Exercises 5.1. It is often said that second language learners have more difficulty acquiring prosody in their second language than most other aspects of speech. Why might it be that prosody is so difficult to master? 5.2. Which word in each pair of rhyming lines of verse contains a syllable that is normally unstressed but has secondary stress and is therefore able to participate in a rhyme? Transcribe this word phonetically, as well at the word it rhymes with. Include primary and secondary stress in your transcription. Note that the use of in these examples follows spelling rules that are no longer current. a. Gleich ist man mit Freudigkeit // Dienstbeflissen und bereit. (“Max und Moritz,” Wilhelm Busch 1970) b. Bringt ihm, was er haben muß: // Zeitung, Pfeife, Fidibus. (“Max und Moritz,” Wilhelm Busch 1970) c. Der alte Narr! Jetzt bleibt er krumm // So äußert sich das Publikum. (“Der alte Narr,” Wilhelm Busch 1958) d. Im Winkel König Fahrenheit // hat still sein Mus gegessen. //– ›Ach Gott, sie war doch schön, die Zeit, // die man nach mir gemessen!‹ (“Kronprätendenten,” Christian Morgenstern 1917) e. Das Perfekt und das Imperfekt // tranken Sekt. (“Unter Zeiten,” Christian Morgenstern 1917) f. Ihr kennt sie, die Leidenschaft, // die uns verbindet: // Helfen, helfen, mit einer Kraft, // die alles überwindet. (“an viele,” Christian Morgenstern 1914) g. War einmal ein Bumerang; // War ein weniges zu lang. (“Bumerang,” Joachim Ringelnatz 1923) h. Wenn ich in die Stube speie, // Lacht mein Bruder wie ein Schwein. // Wenn er lacht, haut meine Schwester, // Wenn sie haut, weint Mütterlein. (“Aus meiner Kinderzeit,” Joachim Ringelnatz 1928) i. Laßt uns im Gleichschritt aufmarschieren, // Ein stolzes Regiment. // Laßt die Fanfaren tremulieren! // Faltet die Fahnen ent! (“Turnermarsch,” Joachim Ringelnatz 1923)

108  Prosody 5.3. Phrasal stress. Use the symbol for primary stress, [ˈ], to indicate which syllable bears the primary stress in each of the following phrases. a. ein sehr schöner Abend b. ohne einen Grund c. ziemlich langsam d. Türen aus Holz e. den Fluss entlang f. mit ihren Mitarbeitern zufrieden g. der großen Kälte wegen h. das Buch auf dem Tisch in der Ecke i. seinem Vater sehr ähnlich j. jungen Menschen gegenüber 5.4. The following sentences all have broad focus and default word order. Identify the elements that bear the nuclear stress, using one of the following labels: subject, accusative object, dative object, verb, time adverbial (when), instrumental adverbial (with what), directional adverbial (where to). Adverbials can be a single word ( ‘yesterday’) or a prepositional phrase ( ‘in the afternoon’). If a noun phrase or a prepositional phrase is the element that bears nuclear stress, remember that it will be the last content word in the phrase that bears this stress. a. Wir sind gestern nach HAUse gekommen. b. Ich wollte mit einem Bohrer ein LOCH in eine Wand bohren. c. Der Kleine schläft gerade. d. Eine Frau sucht nach 50 Jahren ihren SOHN. e. Meine Nachbarin hat ihrem Mann einen MerCEdes geschenkt. f. Ein Polizist hilft einer FRAU über die Straße. 5.5. All the elements you identified in exercises 5.4 can bear the nuclear stress in a sentence with broad focus. If more than one of these occur in the same sentence, one will have precedence over the others (it will be higher in the focus hierarchy than the others). On the basis of the information in exercise 5.4, construct a focus hierarchy using the symbol “>” (where “a > b” means that ‘a is higher on the focus hierarchy than b’). Include all the elements you identified in exercise 5.4 in this hierarchy. 5.6. Predicative expressions (adjective phrases, noun phrases, prepositional phrases) can also be focused in sentences with broad focus. The predicative expressions in (a) through (c) express a property of the subject; those in (d) through (f) express a property of the object. Use the information in these sentences to revise the focus hierarchy you constructed in exercise 5.5 (indicate where a predicative expression fits in the hierarchy).

Prosody  109 a. Die Jungs bleiben gesund. b. Christina will ÄRZtin werden. c. Der Tourismus ist von großer wirtschaftlicher BeDEUtung. d. Ich finde den Geruch von Kaffee einfach HIMMlisch. e. Eine Anwältin hat den Präsidenten einen LÜgner genannt. f. Wir haben ein Mädchen zur KLAssensprecherin gewählt. 5.7. Below are examples of verb phrases in citation form. The citation form of a verb phrase in German uses the infinitive form of the verb, which follows all the other elements in the phrase (noun phrases, prepositional phrases, and so on). You can think of the citation form as the answer to the question . On the basis of the information provided by these examples, formulate a rule that accounts for the placement of primary stress in verb phrases. Hint: Are there similarities between primary stress placement in verb phrases and the placement of nuclear stress in sentences with broad focus? a. nach ˈHause gehen b. einem Freund ein Geˈschenk geben c. einen ˈBrief ins Postfach legen d. mit dem Bus in die ˈStadt fahren e. die ˈZeitung lesen f. einer Perˈson über die Straße helfen g. ˈglücklich sein h. eine Person einen ˈLügner nennen i. ˈschlafen j. ˈarbeiten wollen 5.8. Assume that all of the following sentences have broad focus. Determine which word in each sentence bears the nuclear stress, and underline the stressed syllable in that word. a. Maria hat gestern einen Mercedes gewonnen. b. Ich finde die Schauspieler schlecht. c. Der Junge wollte die Hände in die Tasche stecken. d. Wir wollen Freunde bleiben. e. Mein Kind ist seit gestern krank. f. Karl hat tanzen wollen. g. Ein 17-Jähriger ist nachts aus dem Fenster gestiegen. h. Ein Freund hat mir schon ein Weihnachtspaket geschickt. i. Ein Hausherr hat einen Nachbarn für einen Einbrecher gehalten. j. Peter hat einen Freund ins Haus gebracht. k. Setz dich! l. Sei nicht so laut!

110  Prosody 5.9. Determine whether each of the following responses has broad or narrow focus, based on what is being asked in the question that precedes it. a. Was machst du? Ich sehe einen FILM. b. Welche Farbe hat dein Auto? Es ist ROT. c. Wo wohnt er? In HEssen. d. Was ist gestern passiert? Wir sind ins RestauRANT gegangen. e. Wann kommt Simon vorbei? Um SECHS. f. Was gibt’s Neues? Es SCHNEIT. g. Hast du einen schwarzen Hund oder eine schwarze Katze gesehen? Eine schwarze KAtze. h. Mit wem ist er gefahren? Mit AnDREas. i. Was hast du herausgefunden? Roberts Chef ist ein Betrüger. j. Wer hat das Buch gelesen? MaRIa hat es gelesen. k. Was ist das für ein Krach? Ein Nachbar mäht den RAsen. l. Was gibt’s Neues von Bernd? Er hat ein neues AUto gekauft. 5.10. Narrow focus. In each of the exchanges below, determine which word/phrase in the utterance in B should receive focus, given the question asked in A. Underline the syllable in each focused word/phrase that receives the nuclear stress. a. A: Wann kommt Paul nach Hause? B: Er kommt am Wochenende nach Hause. b. A: Wo haben die Kinder gespielt? B: Sie haben im Garten gespielt. c. A: Was hast du in dem neuen Restaurant gegessen? B: Ich habe eine Pizza gegessen. d. A: Welches Instrument hat der Junge gespielt? B: Er hat Geige gespielt. e. A: Wie hat er das Auto gefahren? B: Er hat es sehr vorsichtig gefahren. f. A: Warum läuft Katja so langsam? B: Weil sie so alt ist. g. A: Was trinken Sie gern zum Frühstück? B: Tee trinke ich gern. h. A: Weißt du, wo mein Spazierstock ist? B: Er ist im Schrank. i. A: Wohin wollte Frau Kell am Wochenende reisen? B: Sie wollte nach Berlin reisen. j. A: Was hat der Student nach der Sprechstunde gelesen? B: Er hat seine E-Mails gelesen.

Prosody  111 5.11. Use a double slash, , to divide the following utterances into intonation phrases. Note that some utterances consist of a single intonation phrase. a. Wir wissen nicht, wann der Schnee aufhört. b. Morgen kommen wir um zwölf vorbei. c. Sechs plus vier mal sieben ist vierunddreißig. d. Katja liebt; Martin nicht. e. Die Studenten lernen in der Bibliothek. f. Lehrer brauchen Kreide, Bücher, Stifte und Papier. g. Er kommt nicht mit, weil er kein Geld hat. h. Auf der linken Seite ist eine Kuh, und auf der rechten Seite ist ein Pferd. i. Eva und Peter telefonieren. j. Oma hat uns nicht gesagt, ob sie selbst fährt. k. Wenn du Lust hast, können wir ins Museum gehen. l. Der Film, den wir heute Abend sehen, heißt Soul Kitchen. m. Fatih Akin, der bei Soul Kitchen Regie führte, lebt in Hamburg. 5.12. Determine whether the intonation contours at the end of the following utterances would be rising or falling in the default case (for example, do not include a contour that would be used to signal politeness). a. Der Kellner sieht fern? b. Was machst du? c. Vielen Dank! d. Moment Mal! e. Wann fängt das Konzert an? f. Der Mantel hat so viel gekostet? g. Um wieviel Uhr bist du heute Morgen aufgestanden? h. Ich muss gestehen, dass ich nicht zu den großen Basketballfans gehöre. i. Hast du Hunger? j. Da mache ich mit! 5.13. How would the interpretation of the utterances below differ, depending on the two types of intonation indicated after each utterance? a. Guten Morgen (level, falling) b. Es ist warm (rising, falling) c. Das Buch ist langweilig (level, rising) d. Mach das nicht (falling, rising) e. Schön (rising, level) f. Hallo (falling, rising) g. Die Kinder waren brav (level, rising) h. Jawohl (level, falling)

112  Prosody 5.14. Read the following ambiguous utterances and determine two possible meanings for each. What are the two meanings for each utterance, and how is prosody used to convey each meaning (how many intonation phrases are there, where are the pauses, and so on)? a. 6 + 3 × 14 b. Tina hilft Julius nicht c. Der Mann folgt der Frau mit dem Motorrad d. Ich esse Brot und Salat oder Suppe e. Es waren alte Männer und Frauen

Chapter 6

Alternations

6.1. Introduction By now you are aware of a number of ways in which sounds alternate in German. The alternation between [ g] in ‘days’ [taːgə] and the [k] in ‘day’ [taːk], which we call final devoicing, is one example of how sounds alternate in different forms of a word. Another example is the alternation between [ g] and [ç] in the suffix , as we see in the two forms of the adjective ‘funny’, one with a vowel-initial ending, , and one without an ending: [lʊstɪgə], [lʊstɪç]. Many alternations are automatic. Native speakers of German automatically devoice obstruents at the end of syllables. They automatically shorten long vowels that are unstressed. They automatically pronounce /ç/ as [x] when it follows a nonfront vowel. They do this without thinking about it, and unless they know something explicitly about the phonetics and phonology of German, they are not aware of what they are doing. These alternations are conditioned by the environment in which a sound occurs (for example, at the end of a syllable, in an unstressed syllable, after a nonfront vowel, and so on). There are other alternations, however, that are not automatically conditioned by a sound’s environment. The vowels represented by , , , and often alternate with their umlauted counterparts—for example, in singular and plural forms of nouns. The plural of ‘harbor’ [haːfən] is ‘harbors’ [heːfən]; the plural of ‘oven’ [ʔoːfən] is ‘ovens’ [ʔøːfən]. Not all nouns form their plural this way, though. The plural of ‘hook’ [haːkən] is ‘hooks’ [haːkən]; the plural form of this word is identical to the singular form. There is nothing about the sounds in words like and that determines whether or not the in them will be umlauted to form the plural. Speakers of German have to learn this in the same way they have to learn if the noun is masculine, feminine, or neuter. In this chapter we will talk about some of the common alternations in German, a number of which you have already encountered in previous chapters. We will first look at alternations that are automatic. Then we will consider those like umlaut, which are not automatic. These nonautomatic alternations are morphophonemic alternations, which involve the substitution of one phoneme for another in a morphological context (in the inflection of words and the creation of new 113

114  Alternations words with derivational affixes). Morphophonemic alternations typically do not affect all words in a language equally, but apply to only a subset of words. We can express all the alternations that we discuss in this chapter as rules. The rules that apply to automatic alternations like final devoicing and vowel shortening, for example, are statements that describe how the phonemes of German are realized phonetically. The rules that describe morphophonemic alternations like umlaut tell us what happens to phonemes in specific morphological contexts. 6.2. Automatic Alternations When we describe automatic alternations like final devoicing, vowel shortening, and so on, we will generally use the following form of notation: (1)

A → B / X _____ Y

In this notation, A stands for a phoneme or a class of phonemes. B expresses the changes that A undergoes as a result of the rule; it indicates how A is realized phonetically. The information to the right of the slash indicates the environment in which the rule operates. The focus bar, _____, indicates the position of the phoneme or phonemes undergoing a change, and X and Y describe the environment that must occur before and/or after A in order for the change to take place. The rule in (1) can be read as “A is realized as B when it occurs between X and Y.” The form of notation in (1) is useful because it allows us to express rules in a very explicit way. We will use some additional symbols and conventions in our discussion of alternations in German. These we will explain as they come up in our discussion. We capitalize the names of the rules that we formulate to describe the alternations in German to distinguish them from terms used to describe general processes that may apply in other languages. 6 .2.1. Vowel Shortening Compare the verb ‘to test’ with the noun it is derived from, ‘test’. Both have the root , which we transcribe phonemically as /pʀoːb/. In the noun , the root vowel is pronounced [oː], but in it is pronounced [o] because it is unstressed (the verbal suffix bears the main stress): [ˈpʁoːbə]; [pʁoˈbiːʁən]. This shortening under absence of stress does not just affect /oː/; it happens to all long vowel phonemes in German, as you learned in chapter 3. We can state this rule, which we call Vowel Shortening, as follows. (2) Vowel Shortening V → [−long] / ___________ [−stress]

The symbol V stands for “vowel,” and “[−long]” and “[−stress]” are features. The rule says that vowel phonemes in German are shortened (they are realized as

Alternations  115 [−long]) when they are in a position in which they are unstressed. If a short vowel phoneme occurs in an unstressed position, it will not be affected by the rule because it is already [−long] (we also say that the rule applies vacuously to short vowels). The root /blɔnd/, for example, has a short vowel, /ɔ/, which is short when stressed in the adjective ‘blond’ as well as when unstressed in the verb ‘to bleach’: [ˈblɔnt]; [blɔnˈdiːʁən]. 6.2.2. Final Devoicing We can express the rule of Final Devoicing by using the feature [sonorant]. Vowels, trills, approximants, and nasals are [+sonorant]; obstruents (plosives and fricatives) are [−sonorant]. (3)

Final Devoicing [−sonorant] → [−voice] / _____ [−sonorant]0.

The subscript zero on the feature [−sonorant] stands for zero or more segments with the feature [−sonorant]. The rule says that obstruents (phonemes that are [−sonorant]) are realized as voiceless when they occur before a syllable boundary (signaled by the period), with zero or more obstruents between them and the syllable boundary. We can demonstrate how this rules works by applying it to the /b/ in different forms of the verb . (4)

a. b. c. d.

‘live!’ ‘(he) lives’ ‘(you) live’ ‘to live’

/leːb/ /leːbt/ /leːbst/ /leː.bən/

→ → → →

[leːp] [leːpt] [leːpst] [leː.bən]

In (4a), /b/ occurs directly before a syllable boundary and is devoiced; zero obstruents occur between it and the syllable boundary. In (4b), one obstruent occurs between /b/ and a syllable boundary, in which case /b/ is devoiced. In (4c), two obstruents occur between /b/ and a syllable boundary and it is again devoiced. Note that in all of these examples, (4a) through (4c), /b/ occurs in a coda. In (4d), on the other hand, Final Devoicing does not apply because /b/ occurs after a syllable boundary; it occurs in an onset. 6.2.3. Dorsal Fricative Assimilation In chapter 3 we took the position that [ç] and [x] are allophones of a single phoneme, and we chose to represent this phoneme with the symbol for the palatal allophone, /ç/. This means that we assume the phoneme to be realized as [ç] except in a few specific environments, where it has a velar pronunciation, [x]. For example, /ç/ is realized as [ç] in words like ‘I’, ‘bad luck’, ‘soft’, and ‘milk’, but as [x] in ‘cake’, ‘hole’, ‘to make’, and ‘smoke’. You know from chapter 3 that /ç/ is pronounced

116  Alternations [x] when it appears after any of the following vowels: /uː ʊ oː ɔ aː a aʊ̯/. When we express this rule formally, using features, we can see clearly why /ç/ takes on a velar pronunciation following this particular set of vowels. We call this rule Dorsal Fricative Assimilation. We use the term “dorsal” because the phoneme /ç/ is [dorsal] and both of its allophones, [ç] and [x], are [dorsal] (both are articulated using the back of the tongue, the dorsum). The rule also involves assimilation, a process in which one sound becomes similar to a neighboring sound. Put simply, this rule tells us that the phoneme /ç/ has a more nonfront pronunciation when it occurs after nonfront vowels. It is pronounced [x], which is [dorsal, −front], rather than [ç], which is [dorsal, +front] ([x] is articulated with the back of the dorsum; [ç] is articulated with the front part of the dorsum): (5) Dorsal Fricative Assimilation /ç/ → [−front] /  V _____ [−front]

By identifying the vowels /uː ʊ oː ɔ aː a/ and /aʊ̯/ as [−front], we can capture the assimilatory nature of this rule. One additional aspect of Dorsal Fricative Assimilation that bears mentioning here is a factor that prevents it from applying. Consider the following words: (6)

a. ‘little cow’ /ˈkuːçən/ [ˈkuːçən] b. ‘female pet owner’ /ˈfʀaʊ̯çən/ [ˈfʁaʊ̯çən] c. ‘biochemistry’ /biːoːçeːˈmiː/ [bioçeˈmiː]

In all of these examples, /ç/ occurs after a nonfront vowel but it is not realized as [x], which is what we would expect, given the rule of Dorsal Fricative Assimilation. What characteristic do these words have that prevents Dorsal Fricative Assimilation from applying to them? In particular, how is ‘little cow’ different from ‘cake’, which has the same phonemic representation, but is pronounced with [x]? (7)

a. ‘cake’ /ˈkuːçən/ [ˈkuːxən] b. ‘little cow’ /ˈkuːçən/ [ˈkuːçən]

You may have noticed that is a complex word, formed with the root and the diminutive suffix , whereas is a simplex word (it does not contain any prefixes or suffixes and is not a compound word). If you analyze the remaining words in (6), you will see that they, too, are complex words. Each of them is composed of smaller, meaningful units, which we call morphemes. (A morpheme is the smallest unit of language that bears meaning.) We can use the plus symbol, which indicates morpheme boundaries, to illustrate the complex nature of the words in (6):

Alternations  117 (8)

/ˈkuː+çən/ [ˈkuː+çən] /ˈfʀaʊ̯+çən/ [ˈfʁaʊ̯+çən] /biːoː+çeːˈmiː/ [bio+çeˈmiː]

Notice that a morpheme boundary precedes /ç/ in each of these words. This is what prevents Dorsal Fricative Assimilation from applying. The morpheme boundary blocks assimilation: /ç/ cannot assimilate to a preceding vowel if a morpheme boundary intervenes. 6.2.4. Spirantization As we mentioned at the beginning of this chapter, [ g] and [ç] alternate in different forms of words with the suffix . (9)

a. ‘calmer’ [ʁuːɪgɐ] b. ‘to calm’ [bəʁuːɪgən]

‘calm’ [ʁuːɪç] ‘calms’ [bəʁuːɪçt]

We assume that the suffix is represented phonemically as /-ɪg/, which means that the phoneme /g/ is sometimes realized as the fricative [ç]. We call the rule that accounts for the fricative pronunciation of /g/ Spirantization (the word spirant is another term for ‘fricative’). (10) Spirantization /g/ → [ç] / /ɪ/ _____ C0. [−stress]

This rule states that /g/ is pronounced as [ç] when it occurs in a coda after an unstressed /ɪ/, with zero or more consonants between it and the following syllable boundary (the symbol C stands for “consonant”). Notice that we have to assume that Spirantization applies before Final Devoicing. We can demonstrate this by working through the derivation of the pronunciation of a word like ‘funny’. (We call the process of applying rules to a phonemic representation to produce the phonetic representation a derivation.) The derivation in (11) shows that the ordering of Spirantization before Final Devoicing produces the desired output. (11) phonemic representation /vɪtsɪg/ vɪtsɪç vac. phonetic representation [vɪtsɪç]

Spirantization Final Devoicing

Final Devoicing applies vacuously (vac.) in (11) because [ç] is voiceless to begin with. The derivation in (12) shows that the opposite ordering of rules produces the nonstandard form [vɪtsɪk].

118  Alternations (12) phonemic representation phonetic representation

/vɪtsɪg/ vɪtsɪk n.a. [vɪtsɪk]

Final Devoicing Spirantization

Spirantization cannot apply in (12) because the structural description of the rule is no longer met at this point in the derivation (there is no longer a /g/). We use the term “nonstandard” because this pronunciation is not accepted as the norm in the standard language. However, it is the norm in colloquial German spoken in southern Germany. We will talk further about this and other examples of regional variation in chapter 7. 6.2.5. Glottal Stop Insertion Because the distribution of the glottal stop in German is predictable, we can use a rule to describe the environments in which it occurs. Recall that glottal stops are found word-initially before a vowel; within a compound before a word that begins with a vowel; and after a prefix if the following sound is a vowel. We can express this rule in a very simple way using the concept of phonological word—a phonological unit (represented by the symbol “ω”) that is larger than the syllable but smaller than the intonation phrase. A phonological word is not the same as a grammatical word, since a grammatical word can consist of more than one phonological word. The compound word ‘veterinarian’, for example, a grammatical word, consists of two phonological words: (Tier)ω (arzt)ω. In German, units that are phonological words are prefixes, roots, and suffixes that begin with a consonant followed by a vowel. Suffixes that do not begin with a consonant followed by a vowel are integrated into the preceding phonological word (Wiese 2000:67). The constituent parts of a compound are also phonological words, as the compound above shows. We can state the rule of Glottal Stop Insertion as in (13), where the symbol Ø stands for “null segment.” (13) Glottal Stop Insertion Ø → [ʔ] / (_____ V . . .)ω

This rule says that a glottal stop is inserted at the beginning of a phonological word before a vowel. It accounts for the presence of the glottal stop in the following examples. (14) a. ([ʔ]Arbeit)ω ‘work’ b. (Haus)ω([ʔ]arbeit)ω ‘homework’ c. ([ʔ]er)ω([ʔ]arbeiten)ω ‘to acquire’

In (14c), for example, a glottal stop is inserted before the prefix and again before , both of which are phonological words that begin with a vowel.

Alternations  119 Although the second syllable of a word like ‘to see’ or ‘chaos’ begins with a vowel, a glottal stop is not inserted before it because there is no phonological word boundary before the vowel. (15) a. (sehen)ω ‘to see’ [zeːən] b. (Chaos)ω ‘chaos’ [kaːɔs]

The suffix in ‘to see’ does not constitute a phonological word because it does not begin with a consonant followed by a vowel. It is part of the phonological word (sehen)ω. 6.2.6. Schwa Deletion The schwa phoneme is unique in German because it is the only vowel phoneme that never occurs in a stressed syllable. Another unique property of schwa is that it can sometimes be deleted. You may have noticed that in the endings /-əm/, /-ən/, and /-əl/, speakers do not always pronounce schwa. Instead, they pronounce the following consonant in such a way that it forms the nucleus of its syllable. That is, when schwa is deleted, the consonant that follows it has a syllabic pronunciation. This is indicated in phonetic transcriptions by a short vertical stroke, [ ]̩  , which is placed beneath the consonant that is syllabic. (16) a. ‘large’ /gʀoːsəm/ [ gʁoːsm̩ ] b. ‘to throw’ /vɛʀfən/ [vɛɐ̯fn̩] c. ‘summit’ /gɪpfəl/ [ gɪpfl ̩]

The deletion of schwa is dependent on the phonetic environment in which schwa appears as well as the level of formality of the speech situation (schwa deletion is more common in informal speech than in formal speech). Capturing all the details of schwa deletion is therefore not an easy task. The major German pronunciation dictionaries, for example, are not always in agreement on the circumstances under which schwa is deleted. To simplify our discussion, we present the data provided in the Duden pronunciation dictionary for “genormte Lautung” ‘standard pronunciation’ (Mangold 2005:37–40). There are various restrictions on the deletion of schwa. One important restriction is that the consonant following schwa has to be in a coda. When the adjective ‘decided, distinct’, for example, occurs without an ending, the /n/ following schwa is in a coda, so schwa can be deleted, as shown in (17a). (17) a. ‘decided, distinct’ /ɛnt.ʃiː.dən/ [ʔɛnt.ʃiː.dn̩] b. ‘decided, distinct’ /ɛnt.ʃiː.də.nə/ [ʔɛnt.ʃiː.də.nə]

When this adjective has an ending, the /n/ following schwa is in an onset, so schwa is not deleted, as we see in (17b).

120  Alternations Another restriction involves the type of consonant following schwa. Schwa cannot be deleted if the consonant is an obstruent. In (18a), schwa is followed by a plosive, so it is not deleted. In (18b) it is followed by a fricative and is also not deleted. (18) a. ‘opens’ /œfnət/ [ʔœfnət] b. ‘beautiful’ /ʃøːnəs/ [ ʃøːnəs]

We can express a simple version of Schwa Deletion as follows: (19) Schwa Deletion /ə/ → Ø / _____ C [+sonorant]

C0.

This rule says that schwa is deleted when it occurs before a sonorant consonant in a coda; zero or more consonants may occur between this sonorant consonant and the following syllable boundary. No consonants follow the sonorant consonant in the examples in (16). In (20a), a single consonant follows; in (20b), two consonants follow. (20) a. ‘binding’ /bɪndənd/ [bɪndn̩t] b. ‘ringing’ /klɪŋəlnd/ [klɪŋl ̩nt]

Notice that Schwa Deletion allows schwa to be deleted when it appears before /ʀ/ (/ʀ/ is a trill, and trills are [+sonorant]). The rules of R-Vocalization and Sonorant Syllabification (see these sections below for details) will ensure that this /ʀ/ is realized as dark schwa, [ɐ]. (21) a. ‘deeper’ /tiːfəʀ/ [tiːfɐ] b. ‘shakes’ /tsɪtəʀt/ [tsɪtɐt]

There are also restrictions on the sounds that precede schwa. We will describe and illustrate these restrictions here, but we will not include them in our rule of Schwa Deletion. We will save the formalization of these restrictions for the exercises at the end of this chapter. The restrictions on the sounds that precede schwa in order for Schwa Deletion to apply are dependent on the consonant that follows schwa. For example, if the consonant following schwa is /m/, then the sound preceding schwa must be a fricative in order for schwa to be deleted. Table 6.1 lists Table 6.1 Restrictions on Schwa Deletion. Phoneme that must precede /ə/ Fricative Obstruent Consonant

When the consonant that follows /ə/ is /m/ /n/ /l/ /ʀ/

Alternations  121 the combinations of sounds that must occur before and after schwa in order for Schwa Deletion to apply. The list starts with the most restrictive environment for Schwa Deletion at the top and ends with the least restrictive environment at the bottom. The most restrictive environment for Schwa Deletion occurs when /m/ follows schwa, since the sound that precedes schwa must be a fricative. (22)

a. b. c. d.

‘deep’ /tiːfəm/ [tiːfm̩ ] ‘stupid’ /bløːdəm/ [bløːdəm] ‘long’ /laŋəm/ [laŋəm] ‘pale’ /faːləm/ [faːləm]

As (22b) shows, schwa cannot be deleted before /m/ if a plosive precedes schwa. It also cannot take place before /m/ if a nasal or lateral approximant precedes schwa, as (22c) and (22d) demonstrate. The following examples show that Schwa Deletion before /n/ is allowed only if the sound preceding schwa is an obstruent (plosive or fricative). (23)

a. b. c. d.

‘store’ /laːdən/ [laːdn̩] ‘pants’ /hoːzən/ [hoːzn̩] ‘long’ /laŋən/ [laŋən] ‘pale’ /faːlən/ [faːlən]

There is an exception to Schwa Deletion before /n/: Although an obstruent precedes schwa in the diminutive suffix , schwa is not deleted: ‘girl’ [meːtçən]. When /l/ follows schwa, the sound preceding schwa must be a consonant in order for Schwa Deletion to apply. (24)

a. b. c. d.

‘poodle’ /puːdəl/ [puːdl ̩] ‘weasel’ /viːzəl/ [viːzl ̩] ‘formula’ /fɔʀməl/ [fɔɐ̯ml ̩] ‘tangle’ /knɔɪ ̯əl/ [knɔɪ ̯əl]

As (24d) shows, Schwa Deletion cannot apply before /l/ if a vowel occurs before schwa. There are no restrictions on the sound that precedes schwa if /ʀ/ follows. As long as /ʀ/ is in a coda, Schwa Deletion will apply. (25)

a. b. c. d.

‘liver’ /leːbəʀ/ [leːbɐ] ‘reader’ /leːzəʀ/ [leːzɐ] ‘bucket’ /aɪ ̯məʀ/ [ʔaɪ ̯mɐ] ‘seer’ /zeːəʀ/ [zeːɐ]

Schwa Deletion applies before /ʀ/ even if a vowel precedes schwa, as (25d) shows.

122  Alternations Schwa Deletion is optional when /m/, /n/, or /l/ follow schwa. A word like , for example, can be pronounced with or without schwa, as [ gʁoːsəm] or [ gʁoːsm̩ ]. Although this rule is optional in the case of /m/, /n/, and /l/, speakers will typically delete schwa unless they are speaking very slowly and clearly. In the case of /ʀ/, however, Schwa Deletion is obligatory. If schwa were not deleted before /ʀ/ (when /ʀ/ occurs in a coda), this would result in the unacceptable sequence *[əɐ̯]. 6.2.7. R-Vocalization As you well know, when /ʀ/ occurs in an onset, it is realized as a consonant, that is, as [r], [ʀ], or [ʁ]. When /ʀ/ occurs in a coda, it is realized as a nonsyllabic vowel: [ɐ̯]. We will give you the opportunity to formalize the consonantal realization of /ʀ/ in the exercises at the end of the chapter and concentrate here on the vocalic realization of /ʀ/, illustrated in (26). (26)

a. b. c. d.

‘animal’ /tiːʀ/ [tiːɐ̯] ‘away’ /fɔʀt/ [fɔɐ̯t] ‘prince’ /fʏʀst/ [fʏɐ̯st] ‘(you) may’ /daʀfst/ [daɐ̯fst]

We can formalize the vocalization of /ʀ/ as follows: (27) R-Vocalization /ʀ/ → [ɐ̯] / _____ C0.

This rule states that /ʀ/ is realized as [ɐ̯] when it occurs in a coda, with zero or more consonants between it and the following syllable boundary. 6.2.8. Sonorant Syllabification In the examples in (26), /ʀ/ is preceded by a number of different vowels, but not by schwa. What happens when /ʀ/ is in a coda and is preceded by schwa? What happens, for example, in a word like ‘camp’ /laː.gəʀ/? Schwa Deletion applies because schwa is followed by /ʀ/ in a coda, and R-Vocalization also applies, yielding [ɐ̯]. We need an additional rule (Sonorant Syllabification) that changes the nonsyllabic dark schwa, [ɐ̯], to a syllabic dark schwa, [ɐ]: (28) phonemic representation phonetic representation

/laː.gəʀ/ laː.gʀ laː.gɐ̯ laː.gɐ [laː.gɐ]

Schwa Deletion R-Vocalization Sonorant Syllabification

We can formalize the rule of Sonorant Syllabification as follows:

Alternations  123 (29) Sonorant Syllabification [+sonorant] → [+syllabic] / .C0 _____C0.

This rule says that a [+sonorant] segment will be realized as syllabic if it occurs in a syllable without a nucleus. Sonorant Syllabification does not apply just in the derivation of [ɐ]. It also produces the syllabic pronunciation of /m/, /n/, and /l/. The rule causes /m/, /n/, and /l/ to be syllabic and thus form the nucleus of a syllable when they find themselves in a syllable without a nucleus because of Schwa Deletion. We can illustrate this for /l/ with the derivation of the pronunciation of the word ‘poodle’. (30) phonemic representation  /puː.dəl/ puː.dl puː.dl ̩ phonetic representation [puːdl ̩]

Schwa Deletion Sonorant Syllabification

When schwa is deleted, the final syllable of has no nucleus. Sonorant Syllabification causes the sonorant segment to be syllabic, thereby providing the syllable with the nucleus that it lost as a result of Schwa Deletion. 6.2.9. Nasal Assimilation Schwa Deletion affects more than just the syllabicity of a following consonant. It also affects the place of articulation of a following consonant. If a plosive precedes a schwa that is deleted before /n/, /n/ takes on the place of articulation of the preceding stop. (31) a. ‘to have’ /haːbən/ [haːbm̩ ] b. ‘lay’ /laːgən/ [laːgŋ̍]

If the plosive is bilabial (/b/ or /p/), as in (31a), /n/ will be realized as a bilabial (syllabic) nasal, [m̩ ]. If the plosive is velar (/g/ or /k/), as in (31b), /n/ will be realized as a velar (syllabic) nasal, [ŋ̍]. Notice that the short vertical stroke that indicates that a segment is [+syllabic] is placed above [ŋ] because the descender on this symbol would interfere with the legibility of the diacritic if it were placed beneath the symbol. We can express Nasal Assimilation in two parts, one for the bilabial pronunciation of the nasal, the other for the velar pronunciation. (32) Nasal Assimilation (preliminary version) a. [+nasal, +syllabic] → [labial] / [−continuant, −sonorant, labial] _____ b. [+nasal, +syllabic] → [velar] / [−continuant, −sonorant, velar] _____

The feature [−continuant] (together with the feature [−sonorant]) distinguishes plosives from fricatives. A sound is [+continuant] if air passes through the ­vocal

124  Alternations tract in a continuous stream, [−continuant] if it does not. Plosives are thus [−continuant], whereas fricatives are [+continuant]. The first part of this rule says that a syllabic nasal has to be [labial] if preceded by a plosive that is [labial]. The second part says that it has to be [velar] if a preceding plosive is velar. We can use a special notation that collapses the two parts of this rule into one and allows us to make the generalization that a syllabic nasal preceded by a plosive has to have the same place of articulation as the plosive. This special notation is called alpha notation and uses the Greek symbol alpha, [α]. (33) Nasal Assimilation [+nasal, +syllabic] → [α place] / [−continuant, −sonorant, α place] _____

When alpha is used with the feature [place], it means that the features for place of articulation must match. 6.3. Morphophonemic Alternations As we pointed out at the beginning of this chapter, morphophonemic alternations involve the substitution of one phoneme for another in a morphological context. The two alternations we will talk about here are called “ablaut” and “umlaut.” Although you may not be familiar with the term “ablaut,” you have already encountered the phenomenon in the process of learning German. You will discover that ablaut also plays a role in English. While you are familiar with the term “umlaut,” you may not know that umlaut can also be found in English. We will not formalize the rules that govern ablaut and umlaut in German. We will focus instead on describing these phenomena in the morphological contexts in which they occur and highlighting the regularities that can be useful to language learners. 6.3.1. Ablaut There are two main classes of verbs in German, weak and strong, which are also referred to as “regular” and “irregular.” (Jacob Grimm, of the Grimm brothers, coined the terms “weak” and “strong.”) Weak verbs form their simple past forms and past participles with the suffix : ‘to love’, ‘loved’, ‘loved’. The majority of verbs in German are weak, and when new verbs enter the language, they are also typically weak: ‘to shop’, ‘shopped’, ‘shopped’. There are far fewer strong verbs in German than weak verbs, but most strong verbs are very common and many belong to the basic vocabulary of the language. According to the Duden grammar, there are roughly 170 simplex strong verbs, that is, strong verbs without additional prefixes (Dudenredaktion 2009b:450). Strong verbs form their principal parts (infinitive, simple past, past participle) by alternating the vowel of the root: ‘to help’, ‘helped’, ‘helped’. This type of vowel alternation is known as “Ablaut” in German, and we

Alternations  125 will use the term “ablaut” here (the term “vowel gradation” is also used). Many of the irregular verbs in English use ablaut. Some common examples are verbs like (, , ), (, , ), and (, , ). While the vowel alternations in these verbs are similar to those in the related (cognate) verbs in German, this is not always the case. For example, compare English (, , ) with German ‘to find’ (, , ). There are a number of different ablaut patterns in German—roughly forty, according to Duden (Dudenredaktion 2009b:452). For example, the verb ‘to ride’ follows the pattern - - ; follows the pattern - - ; and ‘sleep’ follows the pattern - - . Table 6.2 illustrates the ten most common ablaut patterns in German. This table presents the ablaut patterns orthographically, so it does not make a distinction between short and long , , , or , but it does make a distinction between short and long . This means that the ablaut pattern - - , for example, includes verbs with long ( ‘to fly’, ‘flew’, ‘flown’) as well as verbs with short ( ‘flow’, ‘flowed’, ‘flowed’). In spite of the abundance of ablaut patterns, there are some regularities that can be observed (Dudenredaktion 2009b:452–454). • In the simple past of strong verbs, the only vowels that occur are short and long , , , and (where long is written ). This means that you will only find simple past forms like ‘found’, ‘asked’, ‘rode’, ‘remained’, ‘flowed’, ‘offered’, ‘grew’, and ‘hit’. You will never find simple past forms of strong verbs with , , , , or diphthongs. • There are three general ablaut patterns: ABA, ABB, and ABC. That is, the vowel of the simple past will always differ from the vowel of the infinitive, and the vowel of the past participle will either be identical to the vowel of Table 6.2 Ablaut patterns. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Ablaut pattern ei - i - i ei - ie - ie ie - o - o e-o-o i-a-u i-a-o e-a-o a - ie - a a-u-a e-a-e

Example

reiten - ritt - geritten bleiben - blieb - geblieben fließen - floss - geflossen schmelzen - schmolz -geschmolzen singen - sang - gesungen schwimmen - schwamm - geschwommen sterben - starb - gestorben schlafen - schlief - geschlafen wachsen - wuchs - gewachsen geben - gab - gegeben

Meaning

‘to ride’ ‘to remain’ ‘to flow’ ‘to melt’ ‘to sing’ ‘to swim’ ‘to die’ ‘to sleep’ ‘to grow’ ‘to give’

126  Alternations the infinitive, identical to the vowel of the simple past, or different from both. The verbs , , and exemplify these three basic patterns. (34) a. ABB: - - b. ABC: - - c. ABA: - -

We list the ABB pattern first because roughly half the strong verbs in German follow this pattern. There are other regularities that can be observed in the ablaut patterns in German. Rather than simply describe them here, however, we will give you the opportunity to discover them yourself by working with them in the exercises at the end of the chapter. 6.3.2. Umlaut The term “umlaut” can be used in several different ways. You are familiar with the use that refers to the diacritic that is placed over the vowels , , , and in German to represent sounds that are different from those represented by these symbols without the diacritic. “Umlaut” is also used to refer to the replacement of one of these non-umlauted sounds in a word by its umlauted counterpart in a related form of the word. For example, when we form the plural of a word like ‘mother’ and replace the by , we say that we “umlaut” the . Umlaut in this sense is the alternation of vowel sounds in a specific morphological context. A third use of the term involves a type of sound change that took place in the history of German. We will talk about this use of the term (which is also known as “vowel mutation”) when we discuss historical developments in chapter 7. In our discussion here we will focus on the use of umlaut that can be found in various morphological contexts. Umlaut is used in the inflection of nouns, adjectives, and verbs. In particular, umlaut is used in the formation of 1) plural forms of nouns, 2) the comparative and superlative forms of adjectives, 3) second and third person singular present tense forms of strong verbs, and 4) subjunctive II forms. Umlaut also plays a role in 5) derivation: it is sometimes used with a derivational affix to create a new word ( ‘tender’, ‘tenderly’); it is also used on its own in the creation of a new word ( ‘kiss’, ‘to kiss’). Although English does not use umlaut in these morphological contexts in a systematic way, there are a few words in English that do exhibit umlaut. We see it in a handful of plural forms: , ; , ; , ; , ; , . There are comparative and superlative forms of that have umlaut, namely, and . The verb , which was derived from the noun , is an example of umlaut used in derivation. Before we look in more detail at the

Alternations  127 Table 6.3 Umlaut alternations. Vowel alternation /uː/ - /yː/ /ʊ/ - /ʏ/ /oː/ - /øː/ /ɔ/ - /œ/ /aː/ - /eː/ /a/- /ɛ/ /aʊ̯/ - /ɔɪ ̯/

Example

‘brother’ - ‘brothers’ ‘young’ - ‘younger’ ‘oven’ - ‘ovens’ ‘liked’ - ‘would like’ ‘came’ - ‘would come’ ‘man’ - ‘masculine’ ‘to run’ - ‘runs’

use of umlaut in German in the five morphological contexts identified here, we will discuss the phonological relationship between the vowels that participate in this alternation. Table 6.3 lists all the umlaut alternations in German. These correspondences never vary. For example, the only sound that [ʊ] () can alternate with is [ʏ] (typically represented by ); the only sound that [oː] () alternates with is [øː] (); and so on. The examples in (35) demonstrate that this is the case regardless of the morphological category. (35)

[oː] - [øː] a. ‘oven’ /oːfən/ - ‘ovens’ /øːfən/ b. ‘large’ /gʀoːs/ - ‘larger’ /gʀøːsəʀ/ c. ‘to push’ /ʃtoːsən/ - ‘pushes’ /ʃtøːst/ d. ‘bended’ /boːg/ - ‘would bend’ /bøːgə/ e. ‘red’ /ʀoːt/ - ‘to redden’ /ʀøːtən/

How can we describe this alternation in phonological terms? If we look at the first vowel in each pair in table 6.3 (ignoring for now the diphthongs), a feature that all these vowels have in common is [−front]: none of these vowels is a front vowel. The second vowel in each pair, in contrast, has the feature [+front]. In phonological terms, we can say that umlaut in a fronting process that replaces a [−front] vowel with its [+front] counterpart. In the case of the vowels represented by (/aː/ and /a/), the process is a little more complicated. The [−front] counterparts of these vowels would be /æː/ and /æ/, sounds that do not exist in German ([æ] is the sound in English ). The German vowels that are [+front] and articulatorily closest to /æː/ and /æ/ are used instead: /eː/ and /ɛ/. In the case of the diphthong that participates in this alternation (in the alternation between ‘tree’ and ‘trees’, for example), the second member of the diphthong, [ʊ], can be viewed as being replaced by its [+front] (and [−rounded]) counterpart, [ɪ]. When we discuss umlaut in its historical context in chapter 7, you will find out why this alternation between [−front] and [+front] vowels came about.

128  Alternations 6.3.2.1. Noun Plurals The formation of noun plurals is much more complicated in German than it is in English. In English, the suffix is typically used to signal a plural noun (, ; , ; , ). In German, several different suffixes can be used to signal the plural (, , , , ); the vowel of the noun’s root can be umlauted or umlauted together with the addition of a suffix; or a plural form can be created without a suffix or umlaut. There are three different plural types that involve the use of umlaut: (36) a. umlaut + : b. umlaut + : c. umlaut:

‘man’ - ‘men’ ‘stream’ - ‘streams’ ‘apple’ - ‘apples’

If a noun has the plural suffix , it will always have umlaut too, as long as its root vowel is one that can be umlauted. Of course speakers have to know whether a noun takes the suffix, and this is something that cannot be predicted. It is something all speakers, native as well as nonnative, have to learn. The remaining nouns that have umlaut in the plural will add the suffix as long as they do not end in , , or . 6.3.2.2. Comparative and Superlative The comparative form of adjectives in German is formed with the suffix , and the superlative form with the suffix (or ). (37) a. ‘small’ - ‘smaller’ - ‘smallest’ b. ‘wide’ - ‘wider’ - ‘widest’

The vowels of polysyllabic adjectives are not typically umlauted in the comparative or superlative. The comparative forms of ‘thin’ and ‘dark’, for example, are ‘thinner’ and ‘darker’. However, if an adjective is monosyllabic and it has the vowel , , or , this vowel is sometimes umlauted in the comparative and superlative. (38) a. ‘poor’ - ‘poorer’ - ‘poorest’ b. ‘coarse’ - ‘coarser’ - ‘coarsest’ c. ‘young’ - ‘younger’ - ‘youngest’

In most cases, however, this vowel is not umlauted. (39) a. ‘slim’ - ‘slimmer’ - ‘slimmest’ b. ‘proud’ - ‘prouder’ - ‘proudest’ c. ‘colorful’ - ‘more colorful’ - ‘most colorful’

Alternations  129 The presence of umlaut in the comparative and superlative of monosyllabic adjectives cannot be predicted. It is simply information about the adjective that has to be learned. However, the list of adjectives that do exhibit umlaut is not long, and most are very common. 6.3.2.3. Present Tense of Strong Verbs The second and third person singular present tense forms of strong verbs with long or short are formed by umlauting this vowel. (40) a. ‘to drive’, ‘(you) drive’, ‘drives’ b. ‘to let’, ‘(you) let’, ‘lets’

An exception is the verb ‘to create’, which does not form its second and third person present tense forms with the addition of umlaut. While most strong verbs with or in the infinitive do not have umlaut in their second and third person singular present tense forms, the following verbs do: (41) a. ‘to push’, ‘(you) push’, ‘pushes’ b. ‘to run’, ‘(you) run’, ‘runs’ c. ‘to booze’, ‘(you) booze’, ‘boozes’

There is another vowel alternation in second and third person singular present tense forms that we will mention here, namely the alternation between and /: (42) a. ‘to help’, ‘(you) help’, ‘helps’ b. ‘to read’, ‘(you) read’, ‘reads’

While this alternation involves vowels that are not included in the umlaut alternations listed in table 6.3, it does occur in a morphological environment in which we find umlaut. We mention it here because it is an alternation that speakers need to learn—in the same way they need to learn the umlaut alternation—when they acquire the present tense forms of strong verbs in German. 6.3.2.4. Subjunctive II The subjunctive II in German is formed with the stem of the simple past (it is called “subjunctive II” because it is formed with the second principal part of the verb, the simple past form). If a verb is strong and has a simple past stem with the vowel , , or in its root, this vowel is umlauted in the subjunctive II. (43) a. ‘to come’, ‘came’, ‘would come’ b. ‘to shoot’, ‘shot’, ‘would shoot’ c. ‘to drive’, ‘drove’, ‘would drive’

130  Alternations Modal verbs, with the exception of ‘should’ and ‘to want’, form their subjunctive II by umlauting the root vowel of the simple past stem. (44) a.  ‘to be allowed’, ‘was allowed’, ‘would be allowed’ b.  ‘to be able to’, ‘was able to’, ‘would be able to’ c. ‘to have to’, ‘had to’, ‘would have to’

Verbs that belong in the “mixed” class also use umlaut to form the subjunctive II. (Mixed verbs show characteristics of both weak and strong verbs—they use a suffix and ablaut to form the simple past and past participle.) (45) a. ‘to bring’, ‘brought’, ‘would bring’ b. ‘to think’, ‘thought’, ‘would think’

A subset of mixed verbs with in the simple past represents the umlauted vowel in the subjunctive II with rather than (both of which would have the pronunciation [ɛ]): (46) a. ‘to burn’, ‘burned’, ‘would burn’ b. ‘to know’, ‘knew’, ‘would know’

The subjunctive II forms of these verbs are not commonly used, however. Instead, speakers typically use the form of the subjunctive, for example, ‘would burn’. 6.3.2.5. Derivation Umlaut plays an important role in the derivation of new words. Some suffixes are almost always accompanied by umlaut: (47) a. ‘long’, ‘length’ b. ‘bread’, ‘roll’ c. ‘mouse’, ‘little mouse’

Other suffixes are only sometimes (not systematically) accompanied by umlaut: (48) a. b.

‘beard’, ‘bearded’ (but ‘rusty’,  ‘cloudy’) ‘fool’, ‘foolish’ (but ‘scholastic’,   ‘nomadic’)

Verbs are often derived from nouns and adjectives without the use of special affixes, but with umlaut: (49) a. ‘comb’, ‘to comb’ b. ‘short’, ‘to shorten’

Alternations  131 Not all verbs derived in this way exhibit umlaut. Transitive verbs (verbs that have an accusative object) are more likely to have umlaut; intransitive verbs less likely. The verbs in (49), which have umlaut, are transitive. The following, which are intransitive, do not: (50) a. ‘blood’, ‘to bleed’ b. ‘rotten’, ‘to rot’

In some cases two verbs are derived from a noun or adjective, one with umlaut, the other without. The verb with umlaut is transitive; the verb without umlaut is intransitive. (51) a.  ‘steam’, ‘to steam something’, ‘to steam’ b. ‘sick’, ‘to hurt somebody’, ‘to suffer’

In cases like this, umlaut helps to distinguish between the transitive and intransitive verbs derived from the same noun or adjective. 6.4. Advanced Topics 6.4.1. Voice Versus Spread Glottis Traditionally, the difference between the two sets of plosives in German is treated as one of voicing: /b d g/ are voiced; /p t k/ are voiceless (Wängler 1960, Moulton 1962, Wurzel 1970, Hall 1992, Rubach 1990, Wiese 2000). This is the approach that we take in this book. We will refer to the first set of plosives as lenis, and the second set as fortis. The terms “lenis” and “fortis” are traditional terms and will allow us to distinguish the two sets of plosives. An alternative approach takes the position that the distinguishing feature is not [voice] (Lessen Kloeke 1982, Meinhold and Stock 1982, Iverson and Salmons 1995, Jessen and Ringen 2002). According to this approach, both sets of plosives are voiceless—the lenis as well as the fortis. The difference between the two is that the fortis plosives are [spread glottis]. Sounds that are [spread glottis] have a large glottal opening gesture. In plosives, the feature [spread glottis] is often associated with aspiration, a burst of air that follows the release of the plosive. If we provide a transcription of words with fortis plosives that is a bit more narrow than usual (a transcription with a bit more phonetic detail), you will see that these plosives are pronounced with aspiration, indicated with the diacritic [ʰ]. (52) a. ‘passport’ /pas/ [pʰas] b. ‘pond’ /taɪ ̯ç/ [tʰaɪ ̯ç] c. ‘cash register’ /kasə/ [kʰasə]

132  Alternations If you hold the palm of your hand in front of your mouth when you say the words in (52), you will feel the aspiration—a puff of air. Now compare these words with their counterparts with lenis plosives: (53) a. ‘bass’ /bas/ [b̥as] b. ‘dike’ /daɪ ̯ç/ [d̥aɪ ̯ç] c. ‘lane’ /gasə/ [ gå sə]

The narrow transcriptions in (53) show that the lenis plosives in these words are voiceless, indicated with the under-ring or over-ring diacritic. Narrow transcriptions of additional examples show that lenis plosives are voiced only when they occur between voiced segments. (54)

a. b. c. d. e.

‘law’ /gəzɛts/ [ gə̊ zɛts] ‘the law’ /das gəzɛts/ [d̥as g̊əzɛts] ‘criminal law’ /ʃtʀaːfgəzɛts/ [ ʃtʁaːfg̊əzɛts] ‘a law’ /aɪ ̯n gəzɛts/ [ʔaɪ ̯n gəzɛts] ‘school law’ /ʃuːlgəzɛts/ [ ʃuːlgəzɛts]

The /g/ in , for example, is realized as a voiced plosive only when it occurs between two voiced sounds. It is voiced in (54d) and (54e) because it is preceded (as well as followed) by a voiced sound. Although lenis plosives are typically voiced when they occur between two voiced segments, some speakers do not show voicing consistently (Jessen and Ringen 2002:205). If lenis plosives are treated as being phonemically voiceless, voiced pronunciations can be explained easily as being conditioned by the phonetic environment: the voiceless plosive assimilates to (becomes more like) its voiced neighbors. If, on the other hand, lenis plosives are treated as being phonemically voiced, it is difficult to explain the sporadic devoicing. Voiced environments are conducive to voicing, not devoicing. Although it is phonetically more accurate and phonemically more defensible to treat the distinction between lenis and fortis plosives in German as one involving the feature [spread glottis], we present it here as a distinction in voicing. This allows us to simplify our discussion of the phonetic details of the realization of plosives. It also allows us to simplify our discussion of the phonology of German and treat plosives and fricatives together when we talk about final devoicing. If we were to treat the distinction between lenis and fortis plosives in German as one involving [spread glottis], we would need to posit a rule of Fortition for plosives in addition to a rule of Devoicing for fricatives, since lenis plosives are ­realized as fortis plosives when they occur in word-final position (Jessen and

Alternations  133 ­Ringen 2002:212). We see this in the aspiration of the lenis phonemes /b d g/ when they occur in word-final position. (55) a. ‘praise’ /loːb/ [loːpʰ] b. ‘hand’ /hand/ [hantʰ] c. ‘path’ /veːg/ [veːkʰ]

For more details of an analysis that recognizes [spread glottis] as the feature that distinguishes lenis and fortis plosives in German, see Fagan (2009:21–25). 6.4.2. Glottal Stop Wiese (2000:58–60) argues that the concept of the foot, (F), best accounts for the distribution of glottal stops in German. Like the phonological word, the foot is a prosodic constituent. Wiese defines the foot as a sequence of syllables that begins with a stressed syllable and includes all the following syllables up to the next stressed syllable. Wiese (2000:59) formulates his rule of Glottal Stop Insertion as in (56), where the feature [−consonantal] stands for a vowel. (56) Glottal Stop Insertion Insert [ʔ] in: F[ _____ [−consonantal]

This rule states that glottal stops occur in foot-initial position before a vowel. It accounts for glottal stops that are inserted at the beginning of words like ‘breath’, ‘open’, and ‘exercise’. (57) a. ‘breath’ [ˈʔaːtm̩ ] b. ‘open’ [ˈʔɔfn̩] c. ‘exercise’ [ˈʔyːbʊŋ]

It also accounts for the optional insertion of a glottal stop morpheme-internally before a syllable-initial stressed vowel in a word like ‘theater’ and the name (Krech et al. 2009:53; Rues et al. 2009:42). (58) a. ‘theater’ [teˈʔaːtɐ] or [teˈaːtɐ] b. [beˈʔaːtə] or [beˈaːtə]

An analysis based on the foot is not ideal, however. Glottal stops are not inserted before all syllable-initial stressed vowels (Krech et al. 2009:53). (59) a. ‘spy’ [ ʃpiˈoːn] b. ‘deism’ [deˈɪsmʊs] c. ‘meander’ [mɛˈandɐ]

134  Alternations In addition, the glottal stop is inserted before word-initial vowels that do not bear primary stress. (60) a. [ʔ]Anˈtenne ‘antenna’ b. [ʔ]eˈgal ‘all the same’ c. [ʔ]Iˈdee ‘idea’

Wiese (2000:59) claims that words like those in (60) nevertheless consist of two feet: F[I]F[dee]. We will not go into the issues with this approach to the foot structure of words like those in (60), but you can consult Alber (2001:18) if you are interested in a critique of this analysis. Our approach to Glottal Stop Insertion, which is based on the phonological word, accounts for all environments in which glottal stops are inserted— with the exception of the insertion of glottal stops in words like and . However, this pronunciation is presented as optional in Krech et al. (2009), and it is not noted in the Duden pronunciation dictionary (Mangold 2005). In a discussion of standard pronunciation, Rues et al. (2009:41–42) state that vowel sequences in foreign words should not contain a glottal stop (they include words like and among their examples), but they note that glottal stops are often used in these words. Exercises 6.1. Transcribe the following words phonemically and phonetically. Then indicate which of the following rules has been applied: Vowel Shortening, Final Devoicing, Dorsal Fricative Assimilation, Spirantization, Glottal Stop Insertion. Do not include a rule if it applies vacuously. a. b. c. d. e. f. g. h. i. j. k. l. m. n.

Alternations  135 6.2. Schwa Deletion. Replace the question mark to complete the following rules that describe the restrictions on the segments that precede schwa when schwa is deleted before /m/, /n/, and /l/. Two features that will be useful are [sonorant] and [continuant]. a. /ə/ → Ø / ? _____ /m/ C0. b. /ə/ → Ø / ? _____ /n/ C0. c. /ə/ → Ø / ? _____ /l/ C0. 6.3. Schwa Deletion. Transcribe the following words phonetically. Apply Schwa Deletion whenever it can be applied. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. p. q. r. s. t. 6.4. Consonantal realization of /ʀ/. Use the formalization introduced in this chapter to express the rule that /ʀ/ is realized as [r], [ʀ], or [ʁ] when it occurs in an onset. 6.5. The realization of the phoneme /ʀ/. Is /ʀ/ realized as [ʁ], [ɐ̯], or [ɐ]? Transcribe the following words phonemically and phonetically. a. b. c. d.

136  Alternations e. f. g. h. 6.6. Nasal Assimilation. Transcribe the following words phonemically and phonetically. Apply Schwa Deletion whenever it can be applied, and apply Nasal Assimilation when necessary. a. b. c. d. e. f. g. h. i. j. 6.7. Derivation. Provide derivations for the following words, following the format used in (11), (12), (28), and (30) above. Include only the rules that apply to a given word; do not include a rule if it applies vacuously. Be careful to apply the rules in the order in which they must occur to produce the pronunciation that is acceptable in the standard language. a. b. c. d. e. f. g. h. 6.8. Ablaut. Assign the following verbs to one of the ten ablaut patterns in table 6.2. a. b. c. d. e. f. g.

Alternations  137 h. i. j. k. l. m. n. o. p. q. r. s. t. u. v. w. x. y. z. 6.9. Ablaut. Which verb (one in each set) does not belong and why? Make sure to distinguish between short and long , , , and , as this can make a difference in this exercise. a. b. c. d. e. 6.10. Ablaut. The following verbs do not belong to any of the specific ablaut patterns in table 6.2. They do follow one of the general patterns, though (ABB, ABC, ABA). Which of these patterns do they follow? a. b. c. d. e. f. g. h.

138  Alternations 6.11. What two characteristics do the following strong verbs have in common (other than being strong)? a. b. c. d. 6.12. Umlaut: Noun plurals. Provide the plural form of the following nouns and then transcribe them phonetically. Apply Schwa Deletion whenever it can be applied. Note: Not all of the plural forms involve umlaut. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. 6.13. Umlaut: Comparative and superlative. Provide the comparative and superlative forms of the following adjectives and then transcribe them phonetically. Note: Not all of the comparative and superlative forms involve umlaut. Some involve changes in the form of the stem other than umlaut. a. b. c. d. e. f. g. h. i. j. k. l.

Alternations  139 m. n. o. 6.14. Umlaut: Present tense verb forms. Indicate whether the following verbs are weak or strong. Then provide the second and third person singular present tense forms and transcribe them phonetically. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. 6.15. Umlaut: Subjunctive II. Determine whether the following verbs are strong, weak, or mixed. Then provide the third person singular subjunctive II form for each verb, and transcribe it phonetically. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o.

140  Alternations 6.16. Umlaut: Derivation. Create new nouns by combining the following words and suffixes and umlauting the vowel of the root. Provide the meaning of each new word and transcribe it phonetically. a. , b. , c. , d. , e. , f. , g. , h. , i. , j. ,

Chapter 7

Historical Developments and Regional Variation

7.1. Introduction All aspects of language are susceptible to change over time—the sounds of a language, the vocabulary, word order, and so on. The lexicon (vocabulary) is perhaps one of the components in which change is most easily observable. In German, as in English, new words are constantly being created using the building blocks already available in the language (prefixes, roots, suffixes), words are borrowed from other languages, and existing words acquire new meanings. Words can also become obsolete and eventually disappear from the lexicon. The phonology of a language can also undergo change that is observable over time. While sound changes often begin as slight variations in pronunciation in certain phonetic environments, this can result in significant changes in the sound system of a ­language—the addition of brand-new sounds or the loss of old sounds. It can also result in new allophones of existing phonemes. Before we discuss the specific details of the major sound changes that have occurred in the history of the German language, we will present a brief overview of the types of sound change that have shaped the language. 7.2 . Types of Sound Change Sound change can be conditioned or unconditioned. A change that is conditioned occurs in a specific phonetic environment; a change that is unconditioned happens in all environments. An example of a conditioned change is the loss of schwa in word-final position, which occurred in the history of English. Middle English /nɛkə/ ‘neck’, for example, is now /nɛk/. We can express this change, and sound changes in general, by using the same kind of formal notation we use to express the rules that underlie automatic alternations in a language (final devoicing in German, for example): (1)

/ə/ → Ø / ___ #

The rule in (1) says that schwa is lost when it occurs before a word boundary (#). The rule in (2), which has no conditioning environment, says that /uː/ changes into (develops into or is replaced by) /aʊ̯/. This rule is an example of unconditioned change. It is not limited to a specific environment; it happens in all environments. 141

142  Historical Developments and Regional Variation (2)

/uː/ → /aʊ̯/

This change, like the change in (1), took place in the history of English. It affected the /uː/ in words like Old English /huːs/ ‘house’, which is /haʊ̯s/ in Modern English. A very common type of conditioned change is assimilation, which occurs when a sound becomes more like a neighboring sound. Assimilation of consonants can involve a change in the place of articulation, manner of articulation, or voicing. The English word , which was borrowed from Old French ‘to strengthen’, provides an example of assimilation that involves place of articulation. The alveolar nasal /n/ in assimilated to the labial place of articulation of the following /f/ and became the bilabial nasal /m/ in English . An example of assimilation involving manner of articulation is provided by the English word , which was in Early Middle English and then in Middle English. The /n/ in , which assimilated to the preceding /l/, changed from a nasal to an approximant. An example of voicing assimilation is provided by words like , , and . In Old English, when the voiceless fricative /f/ occurred between voiced sounds, it became voiced. We now have an /f/ in the singular form , for example, but a /v/ in the plural , because speakers of Old English changed their pronunciation of ‘wolves’ from [wʊlfas] to [wʊlvas]. Assimilation can also involve vowels, and one type is the historical process we call umlaut (or vowel mutation). In German, this process produced the vowel sounds that are represented orthographically with the diacritic that we refer to as “umlaut,” . Put simply, we can say that the process of umlaut in German involved the assimilation of a [−front] vowel to the [+front] vowel /i/ in a following syllable. For example, Old High German ‘would like’ became in Middle High German (a later stage of the language). The mid, back, rounded vowel became a mid, front, rounded vowel, , by assimilating to the front vowel in the following syllable. The word in Middle High German is also an example of another conditioned sound change, vowel weakening. In Modern German, endings (on nouns, adjectives, verbs, and so on) typically contain the vowel schwa. In Old High German, on the other hand, endings contained “full” vowels (vowels other than schwa). Because these endings were made up of unstressed syllables (less strongly articulated syllables), the vowels in these endings were reduced over time and ultimately became schwa. In the Old High German word , for example, the of the verbal ending brought about umlaut and was subsequently weakened to schwa, yielding in Middle High German. A final example of a conditioned sound change is the devoicing of obstruents in syllable-final position. This is a sound change that characterizes the Middle

Historical Developments and Regional Variation  143 High German period. In Old High German, /d/ in a word like ‘child’, for example, was pronounced [d]. In Middle High German, /d/ was pronounced /t/ when it occurred in syllable-final position. This change was reflected in Middle High German orthography: the singular form was spelled with a , , but the genitive form, which did not have final devoicing, was spelled with a , . Notice that this change did not produce brand-new sounds in German. It simply produced new allophones for existing phonemes. In Old High German, the phoneme /d/ had one allophone, [d]. In the Middle High Period, it had two allophones, [d] and [t]. When a set of sounds changes in a systematic way, this is often referred to as a sound shift. For example, if all the voiced plosives in a language become voiceless, this can be considered a sound shift. The concept of shift can be applied to vowels as well as consonants. The Great English Vowel Shift, for example, refers to a series of changes in the pronunciation of long vowels that took place in England between 1350 and 1700. In German, the term “Lautverschiebung” ‘sound shift’ is used to refer to changes that affected obstruents. As we will see when we discuss these shifts, some involve unconditioned change (they happen regardless of the phonetic environment). Others, while they affect a set of sounds, result in a variety of changes that are dependent on the environment in which the sound occurs. For example, /p/ (which remains /p/ in English) is shifted to /pf/ at the beginning of a word in German (compare and ); after a vowel, it is shifted to /f/ (compare and ). The final two types of sound change we will discuss here are diphthongization and monophthongization. Diphthongization is the change from a monophthong to a diphthong, and monophthongization is the opposite, the change from a diphthong to a monophthong. An example of diphthongization is provided in (2), illustrated by the change from /uː/ in Old English /huːs/ ‘house’ to /aʊ̯/ in Modern English /haʊ̯s/. An example of monophthongization can be found in Southern American English, where /aɪ ̯/ becomes monophthongized to [aː]: , for example, is pronounced [taːm]. 7.3. Sound Changes in the History of German German, like English, is classified as a Germanic language. This means that it belongs to a family (set) of languages that developed from a common ancestor language, Germanic. Germanic is not an attested language (there are no written texts in existence), but linguists have reconstructed it by comparing the languages that make up this family. The Germanic family is traditionally divided into three branches: East Germanic, now extinct, which included Gothic; North Germanic, today consisting of Icelandic, Faroese (the language of the Faroe Islands), Norwegian, Danish, and Swedish; and West Germanic, now made up of German, Dutch, Frisian, English, Afrikaans, and Yiddish.

144  Historical Developments and Regional Variation Germanic is itself a branch of a larger language family, Indo-European. The branches of the Indo-European family that still have living descendants, in addition to Germanic, are Celtic, Italic, Albanian, Greek, Balto-Slavic, Armenian, and Indo-Iranian. Before we discuss the sound changes that have played a role in the historical development of German, we will briefly mention a sound change that affected all the Germanic languages, the First Sound Shift, also known as Grimm’s Law (although first discovered by the Danish scholar Rasmus Rask in 1818, Jacob Grimm provided the definitive description in 1822). This sound change is one of the major innovations in the phonology of Germanic that distinguishes it from all the other branches of Indo-European. It is also the precursor to the Second Sound Shift, one of the most important sound changes in the history of German. The First Sound Shift involved a shift (a series of shifts) in the articulation of Indo-European plosives. In Germanic, the voiceless plosives changed to voiceless fricatives, then the voiced plosives changed to voiceless plosives, and finally the voiced aspirated plosives (voiced plosives followed by a puff of air represented with the symbol /h/) became plain voiced plosives. Table 7.1 illustrates these three stages of the First Sound Shift. The symbol “>” means ‘becomes’ or ‘changes to’; the symbol “ /f/ /b/ > /p/ /bh/ > /b/

Lat. Lith. Sk.

Eng. Eng. Eng.

The Indo-European plosive /p/, for example, was retained in Latin, as the word shows, but shifted to /f/ in Germanic, as the English cognate demonstrates. One exception to the First Sound Shift involved the voiceless plosives when they occurred after /s/. As the following examples show, /p/, /t/, and /k/ were not shifted when they occurred in this position (“Goth.” stands for Gothic). (4)

a. Lat. b. Lat. c. Lat.

Eng. Eng. Goth. ‘to shave’

Table 7.1 The First Sound Shift. Stage I

/p/ > /f/ /t/ > /θ/ /k/ > /x/

Stage II

/b/ > /p/ /d/ > /t/ /g/ > /k/

Stage III

/bh/ > /b/ /dh/ > /d/ /gh/ > /g/

Historical Developments and Regional Variation  145 When we discuss the sound changes that played a role in the history of German, we will see that the Second Sound Shift affected only one set of obstruents created by the First Sound Shift, namely /p/, /t/, and /k/. 7.3.1. Old High German Old High German (OHG) is the earliest stage of the German language (750–1050 AD) for which there are extensive written documents. The term “High” refers to the geographic area in which the language was spoken, the mountainous region in the central and southern portions of the German-speaking territory, and distinguishes it from the “Low German” spoken in the lowlands in the north. High German is further divided into Central German (in the central region) and Upper German (in the southern region). OHG consisted of a number of different dialects; there was no standard or supraregional variety. See figure 7.1 for a map of the current major German dialect areas, which correspond roughly to those of the OHG period. The most important sound change that distinguishes OHG from all other West Germanic languages is the Second Sound Shift, also known as the High German Consonant Shift. The West Germanic voiceless stops, /p/, /t/, and /k/, became the affricates /pf/, /ts/, and /kx/ when they occurred word-initially, following a consonant (as long as this consonant was not /s/), or when geminated (/pp/, /tt/, /kk/). Following /s/, /p t k/ remained unshifted (we find /st/ in German , for example, just as in English ). In all other positions /p/, /t/, and /k/ became the fricatives /f/, /s/, and /x/. Table 7.2 provides examples from English LOW GERMAN

LOW GERMAN

CENTRAL GERMAN

CENTRAL GERMAN

UPPER GERMAN

Figure 7.1 The major Modern ­German dialect areas (adapted from König 2007:230–231).

146  Historical Developments and Regional Variation Table 7.2 The Second Sound Shift. West Germanic

Old High German

/p/

> /pf/

/t/

> /f/ > /ts/

/k/

> /s/ > /kx/

English



(< Old English )



(< Lat. )

> /x/

German





([kxɔts] in Southern Bavaro-Austrian) ( in many Swiss German dialects) ( in many Swiss German dialects)

and Modern German that illustrate this sound change. The English words contain unshifted /p/, /t/, and /k/; their German cognates contain the affricates and fricatives that resulted from the sound shift. Notice that Modern Standard German does not have an affricate that corresponds to OHG /kx/. The affricate [kx] can only be found dialectally, in the southern portion of the German-speaking region. The Second Sound Shift began around 500 AD in the southernmost portion of the German-speaking territory and gradually spread northward until it eventually petered out, ending at the border between High German and Low German. The Second Sound Shift is also called the High German Consonant Shift because the Low German dialects show no trace of this sound change. When we discuss German dialects later in this chapter, you will see that the Second Sound Shift plays an important role in identifying the major dialect areas of German. Two additional changes in the OHG consonant system are the change from /d/ to /t/ and the change from /θ/ to /d/. (5)

a. /d/ > /t/ b. /θ/ > /d/

Eng. Eng.

OHG OHG

These two changes are usually viewed as a chain reaction. After West Germanic /t/ became /ts/ and /s/ in OHG as a result of the Second Sound Shift, /d/ became /t/, and then /θ/ became /d/. As the English examples in (5) show, these changes did not take place in English. The major change in the OHG vowel system was umlaut. This instance of umlaut is referred to as Primary Umlaut, to distinguish it from another instance that

Historical Developments and Regional Variation  147 high mid low

front

central

back

Figure 7.2 Primary Umlaut.

is associated with Middle High German. Primary Umlaut, which took place at the beginning of the OHG period, involved the change from to before an or in a following syllable. We see this change in the present tense of strong verbs, in comparative and superlative forms of adjectives, in plurals of nouns, and in verbs that have a causative meaning. (In the examples here and elsewhere in this chapter, we follow the convention in historical linguistics and use an asterisk to denote a reconstructed form, that is, one that is not preserved in any written documents. NHG stands for “New High German,” the standard language of the most recent period in the history of German.) (6)

Primary Umlaut a. * > OHG ‘he drives’ (= NHG ) b. * > OHG ‘longer’ (= NHG ) c. * > OHG ‘guests’ (= NHG ) d. * > OHG ‘to fell (cause to fall)’ (= NHG )

Primary Umlaut involved both raising and fronting, as shown in figure 7.2: OHG , a low, central vowel, changed to , a mid, front vowel, when , a high, front vowel, occurred in a following syllable. The change from to also happened before (nonsyllabic ). In Modern German, the vowel that resulted from Primary Umlaut is typically represented orthographically by . This practice makes it easy to see the relationship between words that have a common root, for example, words like ‘long’ and ‘longer’, ‘to fall’ and ‘to fell (cause to fall)’, and so on. In some cases, though, the relationship between words that are related historically has become obscure over time, and is used instead of . For example, although ‘parents’ (< OHG ) is related to ‘old’, it is spelled with rather than . 7.3.2. Middle High German A salient feature of Middle High German (MHG), spoken from roughly 1050 to 1350, is the weakening of unstressed syllables. If you compare words from an OHG manuscript with their counterparts in a MHG manuscript, you will find a variety of vowels in unstressed syllables in OHG, but , which represented schwa, in MHG. (7)

Vowel Weakening a. OHG b. OHG

MHG ‘earth’ MHG ‘to have’

148  Historical Developments and Regional Variation c. OHG d. OHG e. OHG

MHG ‘shepherd’ (NHG ) MHG ‘to cry’ MHG ‘(I) take’

As we mentioned above, vowels that were unstressed in OHG were less forcefully articulated and were thus gradually weakened, so that by the time we reach the MHG period, they have been reduced to schwa. Another important change in German that is evident in MHG is Secondary Umlaut. Although the changes brought about by Secondary Umlaut took place during the OHG period, they were not reflected in the orthography until MHG. Like Primary Umlaut, Secondary Umlaut was caused when or occurred in a following syllable. Unlike Primary Umlaut, which involved the raising and fronting of a short vowel, , Secondary Umlaut involved just fronting and affected long as well as short vowels. We demonstrate the fronting nature of Secondary Umlaut in figure 7.3 using the vowels , , and . In (8) we provide examples of Secondary Umlaut with the “normalized” orthography that is used to reproduce the writing in MHG manuscripts (medieval scribes were not consistent in recording an umlauted vowel or in their choice of symbol for indicating its occurrence, so editors of MHG texts have created a conventional orthography for this purpose). (8)

Secondary Umlaut a. > b. > c. > d. > e. > [ yː]

OHG OHG OHG OHG OHG

MHG ‘news’ MHG ‘would like’ MHG ‘beautiful’ MHG ‘over’ MHG ‘houses’

The umlaut symbol used today to indicate umlauted vowels originated in the MHG practice of writing a simplified above the vowels , , and to indicate the new sounds created by Secondary Umlaut. If we fast-forward to Modern German, we see that Secondary Umlaut resulted ultimately in a new set of phonemes in German, the front rounded vowels, /øː œ yː ʏ/. Primary and Secondary Umlaut of and resulted in sounds that merged with sounds from other historical sources, /ɛ/ and /eː/. Today, typically represents the e-sounds that derive historically from umlauted and .

high mid low

front central unrounded rounded unrounded rounded



Figure 7.3 Secondary Umlaut.

back unrounded rounded

Historical Developments and Regional Variation  149 Two important changes in the consonant system of German can be seen in MHG: Final Devoicing and Palatalization. As we pointed out at the beginning of this chapter, Final Devoicing created new allophones for voiced obstruent phonemes. When these phonemes occurred in syllable-final position in MHG, they were voiceless. Because Final Devoicing was represented orthographically in MHG, we can see the results of this change if we compare the nominative singular forms of masculine and neuter nouns that end in obstruents with their genitive forms. (9)

a. MHG ‘dust’, ‘(genitive) dust’ b. MHG ‘child’, ‘(genitive) child’ c. MHG ‘day’, ‘(genitive) days’

The nominative forms contain symbols for voiceless obstruents; the genitive forms contain symbols for their voiced counterparts. Palatalization involved the creation of a new phoneme, /ʃ/, written , which developed from the OHG consonant cluster /sk/: OHG /sk/ > MHG /ʃ/. (10) a. OHG b. OHG

MHG ‘fish’ MHG ‘ship’

In late MHG, /s/ was palatalized (> /ʃ/) when it occurred before other consonants, /p t m n l v/. (11)

a. OHG ‘small’ > NHG b. OHG ‘snow’ > NHG c. OHG ‘to sleep’ > NHG

Palatalization began in the south and spread northward, but it did not reach all the northern dialects. In colloquial German in northern Germany today, speakers retain the old (unpalatalized) pronunciation, [s], before /p/ and /t/: [spɪts] ‘pointed’; [staɪ ̯n] ‘stone’. The palatalized pronunciation is also not represented consistently in the orthography of Modern German, as the words and demonstrate. 7.3.3. Early New High German Early New High German (ENHG) is generally held to be the period between 1350 and 1650. If you look at a text from this period and compare it with a MHG text, the main differences you will observe will be in the vowels, the result of two unconditioned changes: Diphthongization and Monophthongization. Diphthongization changed the MHG long vowels /iː/, /uː/, and /yː/ to the diphthongs /aɪ ̯/, /aʊ̯/, and /ɔɪ ̯/.

150  Historical Developments and Regional Variation (12) a. /iː/ > /aɪ ̯/ b. /uː/ > /aʊ̯/ c. /yː/ > /ɔɪ ̯/

MHG MHG MHG

NHG ‘time’ NHG ‘mouse’ NHG ‘German’

This change began in the Bavaro-Austrian dialect area (the southeast) and spread to include the Central German dialects. The Alemannic dialect area (southwest) did not undergo this change: ‘in Swiss German’ is pronounced [uː]f Schw[iː]zerd[ yː]tsch in Swiss German. Diphthongization is also absent in Low German. Although these new diphthongs are not found in all German dialects, they have become a part of the standard language. A mnemonic for remembering the details of Diphthongization is the MHG phrase , which is in NHG. (The symbol , which is the Latin letter with a hook, is used in normalized MHG to represent the s-sound that developed from /t/ as a result of the Second Sound Shift. This sound was not pronounced like the /s/ that was inherited from Germanic, but eventually the two sounds merged.) Monophthongization changed the MHG diphthongs , , and to the long vowels /iː/ (still spelled ), /uː/, and /yː/. (13) a. > /iː/ b. > /uː/ c. > /yː/

MHG MHG MHG

NHG ‘kind’ NHG ‘book’ NHG ‘sweet’

Monophthongization affected only the Central German dialects. In BavaroAustrian and Alemannic dialects, the old diphthongs remain diphthongal. Low German had other vowels from the beginning, as the Middle Low German words ‘kind’ and ‘good’ demonstrate. Although Monophthongization is only found in the Central German dialects, it is a change that has found its way into the standard language. A mnemonic for remembering Monophthongization is the MHG phrase , which is ‘dear good brothers’ in Modern German. 7.3.4. Summary Table 7.3 summarizes the important changes in the phonology of German that characterize the three major periods of the language prior to NHG. The table shows that in OHG, for example, /p t k/ have shifted to /pf ts kx/ or /f s x/; /d/ has changed to /t/ and /θ/ has changed to /d/; and has undergone umlaut, changing to when or occur in a following syllable. In the exercises that accompany this chapter, you will be given the opportunity to use the information in this table to help you identify the stage of German in a short text (OHG, MHG, or ENHG). Consider, for example, the text in (14), which is given with its NHG equivalent and an English translation.

Historical Developments and Regional Variation  151 Table 7.3 Sound changes in the history of German. Old High German

Second Sound Shift Chain reaction Primary Umlaut Middle High German Vowel weakening Secondary Umlaut Final devoicing Palatalization

Early New High German

/p/ > /pf/ or /f/ /t/ > /ts/ or /s/ /k/ > /kx/ or /x/ /d/ > /t/ /θ/ > /d/ > before or

Eng. , NHG Eng. , NHG Eng. , NHG Eng. , NHG Eng. , NHG * > OHG (NHG )

Unstressed vowels > /ə/ > before or Syllable-final obstruents become voiceless /sk/ > /ʃ/ /s/ > /ʃ/ before /p t m n l w/

OHG > MHG OHG > MHG (NHG ) OHG > MHG

Diphthongization

/iː uː yː/ > /aɪ ̯ aʊ̯ ɔɪ ̯/

Monophthongization

> /iː uː yː/

OHG > MHG OHG > NHG

MHG < mîn niuweȥ hûs> > NHG

MHG > NHG

(14) Der brunne was küele, lûter unde guot. Der Brunnen war kühl, lauter und gut. ‘The spring was cool, pure and good.’

The word provides evidence that this text cannot be OHG, since it exhibits Secondary Umlaut, which is first found in MHG. The words and contain old diphthongs—these diphthongs have not yet undergone Monophthongization—so the text cannot be ENHG. The word also shows that the text cannot be ENHG, since the in it has not yet undergone Diphthongization. The text must therefore be MHG. It is in fact a line from the Nibelungenlied, an epic poem from the MHG period. The sound changes in the history of German that we have presented here can serve as the means for understanding and learning more about the language than just the phonology. Your knowledge of the Second Sound Shift, for example, can help you see the relationship between a newly encountered word in German and its English cognate and help you remember its meaning or even understand it without having to look it up in a dictionary. For example, if you suspect that

152  Historical Developments and Regional Variation ­ erman is related to a word in English and “undo” the Second Sound Shift G by replacing the with a , this yields the word , which provides a major clue about the meaning of . The exercises in this chapter will give you the opportunity to make use of your knowledge about the Second Sound Shift and discover some of the relationships between words in German and their cognates in English. 7.4. Development of German Spelling If you look up the OHG equivalent of ‘to drink’ in one of the standard OHG dictionaries, the Althochdeutsches Wörterbuch (Schützeichel 2006:359), for example, you will find the word spelled over a dozen different ways: , , , , , , , , , . Just as there was no standard language at the time, there was no standard orthography. The great majority of written records of OHG consist of texts produced by scribes in the scriptoria of monasteries from six main scribal centers, each with its own traditions and each found within a different dialect area. The language of these texts is dialectal, often reflecting the dialect of the area, but not always (the monastery in Fulda lies in the Rhine Franconian area, but the language of the texts produced there is primarily East Franconian). Sometimes the language is a mixture of dialects (the result, for example, of a speaker of one dialect copying a manuscript written by someone with a different dialect). One reason we know so much about dialectal variation in OHG is because scribes from different regions transcribed the same text differently. Previously, the only written language had been Latin, so OHG scribes were faced with the difficulty of representing German with the Latin alphabet. They did not always follow the phonemic principle when representing the sounds of German, that is, they did not consistently use one symbol (grapheme) to represent one sound (phoneme). Sounds that were particularly difficult to represent were the affricates that resulted from the Second Sound Shift, and these were not always given spellings that were distinct from those for corresponding fricatives. If you consider the various ways in which the affricate and fricative resulting from a shifted Germanic /t/ were spelled, you will see that was used to represent both sounds (Sonderegger 2003:246). (15) Germanic /t/

OHG affricate , ,

OHG fricative , ,

The writing found in MHG manuscripts is just as varied as that found in OHG works, differing from writer to writer and from region to region. As in OHG times, it was often difficult to represent German sounds with the Latin alphabet. In particular, because there were no symbols in the Latin alphabet for umlauted vowels,

Historical Developments and Regional Variation  153 medieval scribes did not consistently indicate umlaut and they did not always indicate it in a consistent way. For example, the diphthong written in normalized texts (the umlaut of ; compare MHG ‘eye’ and ‘little eye’) could be represented by a combination of a number of different symbols, including the following: , , , , , , , , , and (Michels 1900:29). During the MHG period there was a tendency to pronounce word-initial and intervocalic /f/ with voicing, and it was often spelled with . The past participle meaning ‘found’ was written as well as ; the word for ‘letter’ was written with in the nominative, (with wordfinal /f/), but in the genitive, (with intervocalic /f/). The voiced pronunciation of /f/ did not ultimately take hold, but the use of to represent this sound in word-initial position has often been retained. This explains the use of today to represent /f/ in words like ‘father’ (< OHG ) and ‘from’ (< OHG ). By the time we reach the ENHG period, the inconsistent nature of German spelling is even more noticeable. Waterman (1976:106) calls it “flagrantly haphazard.” There is no single convention for indicating vowel length, and the many ways in which length is indicated during the ENHG period can still be found in Modern German spelling: (16) a. no indication: b. vowel doubling: c. as length symbol: d. as length symbol: e. as length symbol:

‘praise’ ‘moss’ ‘to take’ ‘low’ [foːkt] (German surname)

The decorative use of multiple consonant symbols (“Konsonantenhäufung”) is characteristic of ENHG orthography, yielding spellings like ‘death’, ‘to do’, ‘and’, and ‘words’. The affricate /ts/ is also represented in numerous ways, including , , , , , , and , among many others, although and are the most common. This sound is still represented orthographically in several different ways, , , and : /hɛʀts/ ‘heart’, /zats/ ‘sentence’, /pɪtsa/ ‘pizza’. By the end of the ENHG period, though, relative consistency prevailed in the representation of umlaut. By this time it had also become standard practice to make use of a superscript for this purpose. The period from the end of ENHG (1650) until roughly 1900 can be characterized as a period of standardization. Efforts of language societies (“Sprachgesellschaften”), whose goals included the creation and standardization of a ­German literary language, together with the work of grammarians, philologists, and other scholars and literary figures, resulted in the gradual codification of a ­literary

154  Historical Developments and Regional Variation standard. Not all were in agreement, though, as to what principles to follow in the standardization of orthography. Hieronymus Freyer (1675–1747) and Johann Christoph Adelung (1732–1806), for example, advocated an approach based on pronunciation: “Schreib, wie du sprichst” ‘Write like you speak.’ They also supported spelling that revealed the derivational relationships among words. This required using instead of to spell words like ‘to color’ and ‘to fell’ because of their relationship to the words ‘color’ and ‘to fall’. Notice that this use of also provides historical information, since the sound represented by developed historically from the sound represented by . Jacob Grimm (1785–1863), who supported this use of , favored an approach to orthography that was much more strongly historical. He argued against the use of as a signal of vowel length if it was not historically motivated: ‘song’ < MHG was acceptable, but not ‘much’ < MHG . He also argued against using as a signal of length unless it represented a sound that had once been pronounced. This meant that ‘true’ would be written and would no longer be distinct from the verb form ‘was’. The could remain in ‘to see’, however, because this had originally been pronounced. The word ‘face’, which comes from the same root, has , pronounced [ç], where has . A standard orthography came about eventually following a number of conferences dedicated to orthographic reform. In 1880, German educator Konrad Duden summarized the results of the 1876 conference in his Vollständiges ortho­ graphisches Wörterbuch der deutschen Sprache ‘complete orthographic dictionary of the German language’ (Duden 1880). This was the first edition of what is now the definitive work on German orthography, Duden: Die deutsche Rechtschreibung, the twenty-fifth edition of which appeared in 2009 (Dudenredaktion 2009a). This edition is based on the German orthographic reform (Rechtschreibreform) of 1996, the goal of which was to simplify the spelling of German. One set of changes brought about by this reform was the use of instead of and instead of in words for which there was a basic form (“Grundform”) with or . (17) a. ‘stem’ instead of ; basic form ‘pole’ b.  ‘effusive’ instead of ; basic form ‘exuberance’ (18) a.  ‘to blow one’s nose’ instead of ; basic form ‘snout’ b.  ‘to bash up’ instead of ; basic form ‘bruised’

Historical Developments and Regional Variation  155 Notice that these changes affected only a handful of words and were not carried out across the board (one could argue that the “Grundform” of ‘parents’ is ‘old’, but this word retained in the root). One change that was carried out without exception and affected numerous words throughout the lexicon was the change governing the use of and to represent /s/. Prior to the 1996 reform, /s/ was spelled with when it appeared intervocalically after a short vowel; otherwise it was written with (intervocalically after a long vowel or diphthong; at the end of a word; before a consonant). This meant that one form of a word could be spelled with , but another form with . (19) Old (pre-1996) spelling a. /dɛs flʊsəs/ ‘of the river’, but /flʊs/ ‘river’ b. /kʏsən/ ‘to kiss’, but /eːʀ kʏst/ ‘he kisses’

It also meant that if appeared at the end of a word or before a consonant, the orthography could not be used to determine whether the preceding vowel was long or short. (20) Old (pre-1996) spelling a. long vowel: /floːs/ ‘raft’ b. short vowel: /ʃpʀɔs/ ‘shoot’

Currently, with the 1996 reforms in place, can always be used to determine the pronunciation of a preceding vowel: /s/ is now typically spelled with after short vowels and with after long vowels and diphthongs. (21)

Current spelling: after short vowels a. /flʊs/ ‘river’ b. /ʃpʀɔs/ ‘shoot’ c. /has/ ‘hate’

(22)

Current spelling: after long vowels and diphthongs a. /fuːs/ ‘foot’ b. /floːs/ ‘raft’ c. /aːs/ ‘ate’ d. /vaɪ ̯s/ ‘white’

This change in the use of and was motivated by the stem principle (“Stammprinzip”) of orthographic representation, which says that a stem should change as little as possible in all words in which it appears. Applied to words that used to be written with in some forms and in others, it now ensures a uniform spelling with : ‘river’, ‘of the river’.

156  Historical Developments and Regional Variation The stem principle is also responsible for lifting the ban against three identical consonant symbols in compound words: the earlier form ‘shipping’ is now written , allowing ‘ship’ to be represented consistently with . This principle is also responsible for putting an back in words like ‘roughness’, which used to be spelled . This very brief overview of the development of German orthography has provided background information for some of the many inconsistencies that remain in current orthographic conventions. For example, the use of and (two different symbols) to represent the same sound, the phoneme /f/, can be traced back to the MHG period when word-initial /f/ tended to be pronounced as a voiced sound and spelled with . The in ‘father’ is no longer a voiced sound, but it retains the symbol . Note that orthographic inconsistency—inconsistency in the one-to-one correspondence between sound and orthographic symbol—is not necessarily undesirable. The use of instead of to represent some instances of /eː/ and /ɛ/ provides readers with an orthographic connection between words that are related inflectionally and derivationally. (23) Inflectional relationship a. ‘to drive’ b. ‘father’ c. ‘long’

‘she drives’ ‘fathers’ ‘longer’

(24) Derivational relationship a. ‘poor’ b. ‘cold’ c. ‘brook’

‘shabby’ ‘cold’ ‘brooklet’

Because there are multiple ways of representing vowel sounds in German, it is also possible to distinguish homonyms orthographically. (25) a. b. c. d. e.

‘song’ ‘string’ ‘to grind’ ‘boats’ ‘honor’

‘eyelid’ ‘side’ ‘to paint’ ‘messenger’ ‘ear’

The current state of German orthography, codified in the Duden spelling dictionary (Dudenredaktion 2009a), reflects a long history of attempts to standardize, simplify, and systematize the way in which the language is represented graphically. It is the product of a number of different principles that have been used to map sounds onto symbols. We see a phonemic principle at work when is used for the [ç] in [køːnɪç] ‘king’ and the [ g] in [køːnɪgə]

Historical Developments and Regional Variation  157 (these sounds are allophones of the phoneme /g/: /køːnɪg/, /køːnɪgə/). The stem principle lies behind the most recent change in the use of and . The homonymy principle (words that are pronounced the same but have different, unrelated meanings should be spelled differently) is responsible for spellings like ‘to empty’ and ‘to teach’. The orthography of German also retains traces of the history of the language itself—from old diphthongs that are now monophthongs but still written with two vowel symbols ( = /iː/ in words like ‘to love’) to the inconsistent representation of the affricate /ts/ ( ‘pain’, ‘net’), evidence of the difficulty faced by OHG scribes when attempting to represent one of the new sounds in OHG brought about by the Second Sound Shift. 7.5. Regional Variation When we talk about regional variation in German, it is useful to identify a continuum of language varieties, the standard-colloquial-dialect continuum. At the one end of this continuum is the standard language, Standard German, the supraregional variety codified in works such as the Duden grammar (Dudenredaktion 2009b), spelling dictionary (Dudenredaktion 2009a), and pronunciation dictionary (Mangold 2005). The pronunciation guidelines we present in this book are those of the standard language. At the other end of the continuum are the dialects, the local or regional varieties. Between these two extremes are the colloquial varieties (“Umgangssprachen”), which are also regional, but associated with much larger geographic regions than dialects. Although Standard German is relatively uniform in grammar, spelling, and pronunciation, it does exhibit regional variation, including regional variation in pronunciation. For example, the different consonantal r-sounds that are acceptable in the standard can be described in regional terms. Uvular r is used predominantly in the north, central, and southwestern parts of the German-speaking region. Alveolar r is used in the southeastern portion (Bavaria and Austria). Greater regional variation in pronunciation is found in colloquial varieties of German as well as in German dialects. In the following section we will describe some of the more salient characteristics of pronunciation in the colloquial varieties of German. Then we will turn to the German dialects, making use of some of the major sound changes in the history of the language to identify the main dialect areas of German. 7.5.1. Colloquial German We will focus here on the variation in pronunciation in the colloquial language spoken in Germany. We can divide Germany into three broad geographic regions that are useful for describing variation in the colloquial language (Durrell 1992:10).

158  Historical Developments and Regional Variation (26) North: north of a line running from west to east through Düsseldorf Central: the central area between North and South South: south of a line running from west to east through Frankfurt am Main

These regions correspond roughly to those that make up the major German dialect areas (see figure 7.1). In northern Germany, the phoneme /ŋ/ is pronounced [ŋk] in word-final position. This means that a word like ‘apartment’ is pronounced [voːnʊŋk] and ‘thing’ is pronounced [dɪŋk]. Another northern German pronunciation is the substitution of [s] for [ ʃ ] syllable-initially before /p/ and /t/: ‘to speak’ is pronounced [spʁɛçən] and ‘piece’ is pronounced [stʏk]. Remember that the standard pronunciation of these words is the result of Palatalization, a sound change associated with the MHG period. As we mentioned earlier, this change began in the south and spread northward but ended before it reached the northern dialects. The colloquial German pronunciation in northern Germany reflects the original (unpalatalized) pronunciation. A final strictly northern German pronunciation that we will mention is the substitution of [ j] for syllable-initial [ g], which is particularly characteristic for colloquial German in Cologne and Berlin: ‘made’ [ jəmaxt], ‘morning’ [mɔɐ̯jən]. The northern and central regions share some pronunciation features. In both regions, syllable-final /g/ is pronounced [x] (after nonfront vowels) or [ç] (after all other sounds): ‘day’ [taːx], ‘path’ [veːç], ‘mountain’ [bɛɐ̯ç]. Another shared feature involves the pronunciation of long . In the standard language, both [ɛː] and [eː] are acceptable pronunciations, although dictionaries typically just give the pronunciation with [ɛː]: ‘beaks’ [ ʃnɛːbəl], ‘namely’ [nɛːmlɪç]. In northern and central Germany, speakers typically use [eː] for this sound: [ ʃneːbəl], [neːmlɪç]. This is the pronunciation we use and advocate here, in part because it is the norm in the colloquial language in a large region of Germany. A pronunciation found in central German is the substitution of [ ʃ ] for [ç]. This means that you will hear speakers from this region say [ʔɪʃ] for ‘I’, [ʁaɪ ̯ʃ] for ‘rich’, and so on. In the south, the suffix is pronounced [ɪk] when it occurs in syllable-­final position: ‘important’ [vɪçtɪk], ‘insulted’ [bəlaɪ ̯dɪkt]. This means that speakers who pronounce this suffix with a [k] do not have the rule of Spirantization. They pronounce /g/ as a [k] when it occurs in syllable-final position (because of the rule of Final Devoicing) and [ g] elsewhere. Another feature of speech in southern Germany is the substitution of [k] for [ç] in word-initial position: ‘chemistry’ [kemiː], ‘China’ [kiːna]. In southwestern Germany, and are pronounced [ ʃp] and [ ʃt] when they occur wordinternally, that is, in environments in which they are pronounced [sp] and [st] in

Historical Developments and Regional Variation  159 the standard language. This means that while a word like ‘constant’ is pronounced [kɔnstant] in the standard language, you will hear it pronounced [kɔnʃtant] in the southwest. Although the [eː] pronunciation of long found in northern and central Germany is fully acceptable in the standard language (and a pronunciation that we recommend), the others presented here are not. For example, the Duden volume that deals with “correct” usage, Duden: Richtiges und gutes Deutsch, labels the substitution of [ j] for [ g] in words like ‘made’ and ‘morning’ as “nicht korrekt” (Eisenberg and Schneider 2011:132). This work is less critical about the substitution of [k] for [ç] in words like , saying that although the pronunciation [kiːna] is common in the south, it is not standard (Eisenberg and Schneider 2011:134). We recommend that you aim for the standard pronunciation in all cases, as you will avoid any stigma that might be attached to a pronunciation that is not standard. 7.5.2. German Dialects The Modern German dialects are typically divided into three major dialect areas, Low German, Central German, and Upper German, with Central and Upper German making up the High German area (see figure 7.1). These three areas can be identified by the extent to which three sound changes in the history of German affected the language spoken in that area. The Second Sound Shift is the most important sound change used in the classification of German dialects. The other two that play a role are Diphthongization and Monophthongization, two sound changes that characterize the ENHG period. When we present dialect data, we will use orthographic symbols as much as possible in order to keep the discussion simple. When necessary, we will also use phonetic symbols. The Second Sound Shift began in the south and gradually spread northward until what is now called the Benrath Line, which runs from west to east through Benrath (a district of Düsseldorf), where it crosses the Rhine, to eastern Germany near Frankfurt (Oder). The Benrath Line is also known as the maken-machen line or isogloss. An isogloss is a line drawn on a dialect map that separates the area where one linguistic form or feature is used from another area with a different form or feature. North of the Benrath Line, West Germanic /k/ in word-internal position remains unshifted, as represented by the word ‘to make’; south of this line, word-internal /k/ shifted to a fricative, , represented by . See figure 7.4 for this and other isoglosses that play a role in delineating the major German dialects. Another important isogloss is the appel-apfel isogloss, or the Germersheim Line (this line crosses the Rhine at Germersheim). North of this line, dialect speakers have an unshifted /pp/ in words like ‘apple’; south of this line, /pp/ has been shifted to , as in . These two isoglosses are important because they delineate the three main dialect areas: North

160  Historical Developments and Regional Variation

Hamburg

Ems

Berlin se r

We

e Elb

maken machen

Pund Fund

Dorp Dorf

Frankfurt Main

r Ege

Do nau

Mosel

Wien

In n

München Basel

Prague

Apfel

Rh ein

dat das

Appel

Figure 7.4 Isoglosses of the Second Sound Shift (adapted from König 2007:64).

of the maken-machen isogloss are the Low German dialects; south of the appel-apfel isogloss are the Upper German dialects; between the two are the Central German dialects. In the Low German dialects, we find no trace of the Second Sound Shift; /p t k/ remain unshifted. In the Upper German dialects, /p t k/ underwent the full effects of the Second Sound Shift, and we even find an affricate, [kx], where Standard German has [k]: South Bavaro-Austrian [kx]alb ‘calf ’ corresponds to Standard German [k]alb. In the Central German dialects we find shifted and unshifted /p t k/. Intervocalically, /p t k/ are shifted to fricatives in all Central German dialects. (27) Intervocalic /p t k/ in Central German dialects ‘to sleep’ ‘water’ ‘to make’

Intervocalic /pp/, on the other hand, remains unshifted in the entire region: we only find ‘apple’. In West Central German dialects, word-initial /p/ remains unshifted; in East Central German dialects it is shifted to .

Historical Developments and Regional Variation  161 (28) Initial /p/ in Central German dialects West Central German: ‘pound’ East Central German: ‘pound’

There is further variation in the West Central German area. In Rhine Franconian (a dialect in the southern portion of this region), we find ‘village’ (with /p/ shifted to after ) and ‘that’ (with /t/ in word-final position shifted to ). In Ripuarian (in the northern portion of this region, in the area around Cologne), these words have unshifted /p/ and /t/: , . The Central German region is a transitional area: in general, the further northward and westward you go, the fewer instances you find of shifted /p t k/. Diphthongization is a change that is of particular importance for the classification of the Upper German dialects. This dialect region is typically divided into three areas: Alemannic, Upper Franconian, and Bavaro-Austrian. We will concentrate here on two of these areas, Alemannic and Bavaro-Austrian. In general, the Alemannic dialects (which include the Swiss German dialects) did not undergo Diphthongization. Recall that the phrase is pronounced [uː]f Schw[iː]zerd[ yː]tsch in the Swiss German dialects. The MHG monophthongs /iː uː yː/ remain monophthongs in the Alemannic dialect area. In contrast, these monophthongs become diphthongs in the dialects in the Bavaro-Austrian region (spoken primarily in southern Germany and Austria). They also become diphthongs in the Central German dialects. In the Low German dialects, however, they remain monophthongs. In Low German we find monophthongs in words like [miːn], ‘my’, [huːs] ‘house’, and [lyːt] ‘people’, where Standard German has diphthongs (, , ). Monophthongization affected only the Central German dialects. The MHG diphthongs , , and became the monophthongs /iː uː yː/ in these dialects ( ‘dear good brothers’), but not in the Upper German dialects, where they remained diphthongs. For example, in Central BavaroAustrian (spoken in the Austrian state of Upper Austria), the words for ‘dear’ and ‘good’ (< MHG ) are and . Monophthongization is moot when it comes to Low German, since Middle Low German (MLG) had monophthongs that corresponded to the MHG diphthongs. Where MHG had and , MLG had and : MHG ‘dear’, MLG ; MHG ‘good’, MLG . Table 7.4 summarizes the role that the Second Sound Shift, Diphthongization, and Monophthongization play in the major German dialect areas. If you are given a sample of a dialect from one of these regions, you can apply this information to determine the region it comes from—as long as the sample has sufficient relevant information. Consider the following sample and look for evidence of these three sound changes:

162  Historical Developments and Regional Variation Table 7.4 Sound changes in the Modern German dialects. Low German Central German West Upper German East Upper German

Second Sound Shift No Yes and no Yes Yes

Diphthongization No Yes No Yes

Monophthongization NA Yes No No

(29) Dialect: Dat beste iss ümmer, sick an de eegn Nääs to fatn und anner Lüh in Ruh to latn! Standard German: Das Beste ist immer, sich an die eigene Nase zu fassen und andere Leute in Ruhe zu lassen! Translation: ‘It’s always best to think about your own mistakes and leave other people in peace.’

There is no evidence in this sample for the Second Sound Shift. The words ‘that’, ‘to grab’, and ‘to let’ have (unshifted /t/) where Standard German has or : , , and . The reflexive pronoun, ‘oneself ’, has unshifted /k/, where Standard German has : . If we look for evidence of Diphthongization, we find the word ‘people’, with a long monophthong, which corresponds to Standard German , with a diphthong. The absence of evidence for both the Second Sound Shift and Diphthongization leads to the conclusion that this must be a Low German dialect. It is in fact an example of Plattdeutsch ‘Low German’ from a Plattdeutsch dictionary on the Norddeutscher Rundfunk (North German Broadcasting) Web site (www.ndr.de). The following sample comes from another dialect area: (30) Dialect: Uffgaabe mache, d Buebe hei chönne öppis baschtle, d Meitli hei chönne lisme. Standard German: Aufgaben machen, die Buben haben etwas basteln, die Mädchen stricken können. Translation: ‘Do homework, the boys could do handicrafts, the girls could knit.’

This is clearly an Upper German dialect. Not only is /k/ shifted intervocalically in ‘to make’, it is also shifted word-initially in ‘to be able to’, where Standard German has with unshifted /k/. The word ‘boys’ confirms this, since it has a diphthong where Standard German has a monophthong (< MHG ), which means that Monophthongization has not occurred in this dialect. The monophthong in ‘homework’, which corresponds to the diphthong in Standard German , demonstrates that Diphthongization is also absent, leading us to conclude that this must be an Alemannic dialect (from the western portion of the Upper

Historical Developments and Regional Variation  163 German dialect area). The sample is in fact an example of Swiss German spoken in Bern (Siebenhaar 2002). In the exercises for this chapter, you will be given the opportunity to identify the dialect area of additional samples of German dialects. 7.6. Advanced Topics 7.6.1. Germanic Accent Shift Scholars are in general agreement that Proto-Indo-European (PIE), the ancestor of the Indo-European languages, was a pitch-accent language (Fortson 2010:68). This means that the accented syllable in PIE had a different pitch (a higher pitch) than neighboring syllables. The accent was also mobile. It was not fixed on a particular syllable, for example, the first or the last. It could occur on different syllables in a word and could thus be used to distinguish meaning. We have some examples of this in English where differences in stress placement are used to distinguish nouns and verbs: is a noun, but is a verb. In Germanic, the pitch accent became a stress accent, with accented syllables differing from unaccented syllables in length (they lasted longer) and loudness (they were pronounced with greater intensity). The accent also became fixed on the initial syllable. Greek ‘father’ is in German and in English. Certain prefixes remained unstressed, as in words like ‘to receive’ and ‘to forget’ in German and and in English. The change to a stress accent led to further changes in the Germanic languages over time. In German, vowels in unstressed syllables were articulated with less intensity and duration, leading to the vowel weakening found in MHG, which we discussed above. For example, the full vowel in the infinitive ending of a group of verbs in OHG became schwa in MHG: OHG ‘to thank’ became MHG . Vowels in unstressed syllables were also lost. OHG is now ‘mercy’; OHG has become ‘church’. The deletion of schwa found in German today in words like ‘to give’ [ geːbm̩ ] and ‘bird’ [foːgl ̩] can be viewed as a current example of change brought on by the shift from a pitch accent to a stress accent. 7.6.2. Verner’s Law There were some exceptions to Grimm’s Law (the First Sound Shift) that Jacob Grimm could not explain. For example, Grimm’s Law predicts that Indo-­European /t/ will become /θ/ in Germanic, and we know that through an additional change it will become /d/ in German. We see these changes in the word ‘brother’, where /d/ corresponds to Indo-European /t/ in Sk. ‘brother’. In the word for ‘father’, however, which also contained a /t/ in Indo-European, as demonstrated by Lat. , German has /t/: .

164  Historical Developments and Regional Variation Because the expected and unexpected reflexes of Indo-European /p t k/ could often be seen in different forms of a verb, like ‘to cut’ and ‘cut’, Jacob Grimm referred to this alternation of consonants as “grammatischer Wechsel” ‘grammatical alternation’. Karl Verner, a Danish linguist, discovered an explanation for the exceptions that led to this alternation—an explanation that we now call Verner’s Law (see Verner 1877): When Indo-European /p t k/ occurred word-internally and were not immediately preceded by a stressed ­syllable, they became the voiced fricatives /β ð ɣ/ in Germanic. Otherwise they became /f θ v/. In reference works on the history of German, you will often see the symbols used for /β ð ɣ/ (/β/ is bilabial; /ð/ is interdental; /ɣ/ is velar). The sound changes brought about by Grimm’s Law and Verner’s Law took place before the mobile accent became fixed on initial syllables. Once the accent became fixed, the environment for the changes brought about by Verner’s Law disappeared. However, Verner was able to discover that there was a correlation between the mobile accent in Indo-European and the alternation of voiced and voiceless fricatives in Germanic. The alternation between /d/ and /t/ in and , for example, could be traced back to differences in the placement of the accent in these words when the accent was still mobile. Indo-European /s/ was also affected by Verner’s Law, which led to alternations between /s/ and /z/ in Germanic. This /z/ became /r/ in West Germanic, a process called rhotacism. The alternation between /s/ and /r/ can be seen in English as well as German. We see it, for example, in the difference in the past tense forms of the verb in English, and . Sometimes a word in English will have an where German has an , as in English and German ‘to lose’. Other times an in English will correspond to an in German, as in English and German ‘rabbit’. Table 7.5 provides examples of how Verner’s Law is realized today in German. The alternations that are due to Verner’s Law can often be seen in words that are related derivationally (in nouns that are derived from verbs, for example), not just in different forms of a verb. If you keep these alternations in mind, you will probably discover relationships between words that you had not been aware of previously.

Table 7.5 Alternations in Modern Standard German resulting from Verner’s Law. Indo-European

/p/ /t/ /k/ /s/

Germanic

/f/ - /β/ / θ/ - /ð/ /x/ - /ɣ/ /s/ - /z/

Modern German

- - - -

Examples

‘yeast’ ‘to cut’ ‘to pull’ ‘being’

‘to raise’ ‘cut’ ‘pulled’ ‘was’

Historical Developments and Regional Variation  165 Exercises 7.1. Sound change. Identify the type of sound change represented by each example: assimilation, diphthongization, final devoicing, monophthongization, palatalization, umlaut (a specific kind of assimilation), vowel lengthening, vowel shortening, and vowel weakening. Pay attention to the segments that are in bold. a. OHG > MHG ‘door’ b. OHG > MHG [ə] ‘to give’ c. OHG > NHG ‘good’ d. MHG [a] > NHG ‘day (genitive)’ e. MHG > NHG ‘my’ f. German > colloquial German [ŋg] ‘unhappily’ g. MHG [aː] > NHG [a] ‘thought’ h. OHG > MHG ‘child’ i. OHG > Late MHG ‘snake’ j. OHG > Late MHG [ə] ‘snake’ k. OHG > MHG ‘cheese’ l. OHG > OHG ‘to name’ 7.2. Second Sound Shift and other consonantal changes. Supply the missing consonants in the German cognates, using orthographic symbols. Then supply the meaning of the German word. As you will see, some of the meanings of these cognates have diverged over time. a.

b.

c.

d.

e.

f. g.

h.

i.

j. k.

l.

m.

n.

o.

p.

q.

r.

s.

t.

166  Historical Developments and Regional Variation 7.3. MHG. The following text is the first two verses from the poem “Under der linden” by Walter von der Vogelweide (1843), considered the greatest lyric poet of the MHG period. Provide evidence from the text itself that the language is MHG. Consult table 7.3 to determine what kind of evidence to look for. You will want to include evidence that shows that this cannot be OHG or ENHG. Under der linden an der heide, dâ unser zweier bette was, dâ mugent ir vinden schône beide gebrochen bluomen unde gras. vor dem walde in einem tal, tandaradei, schône sanc diu nahtegal. Ich kam gegangen zuo der ouwe: dô was mîn friedel komen ê. dâ wart ich enpfangen hêre frouwe, daz ich bin sælic iemer mê. kuster mich? wol tûsentstunt: tandaradei, seht wie rôt mir ist der munt.

Notes: ‘können’; ‘finden’; ‘schöne’ (this is an adverb; is the adjective); ‘Nachtigall’; ‘Au’; ‘Geliebter’; ‘eher; früher’; ‘wurde’; ‘hoch, vornehm’; ‘Dame’; ‘glücklich’; ‘küsste er’; ‘tausendmal’ 7.4. OHG, MHG, or ENHG? Identify the language of each of the following texts as OHG, MHG, or ENHG, providing evidence in support of your conclusion. Consult table 7.3 for the features to search for in these texts. Although a circumflex () is typically used to represent long vowels in MHG (, for example), we use the macron () in all texts, regardless of the language, for any diacritic used to indicate vowel length (, , , and so on). a. denn daß ich meine Leut vberredet/ es were nur noch vmb 3. oder 4. tag zu thun/ darumb solten sie guts muts seyn ( ‘um’; ‘gutes’) b. dā wart gevohten genuoc, der eine stach, der ander sluoc ( ‘gefochten’; ‘genug’) c. dīu hēizza súnna. hāt mír mīne scône benóman. ( ‘heiße’; ‘Sonne’; ‘Schönheit’)

Historical Developments and Regional Variation  167 d. Fuor er ze demo ringe, ze hevīgemo dinge ( ‘fuhr’; ‘Versamm­ lung’; ‘bedeutend’; ‘Gerichtsversammlung’) e. ē daz ich lāze ūz mīner hant/ mīn lēhen und mīne ēre ( ‘ehe’; ‘Lehen’; ‘Ehre’) f. Jch wil deinen Namen predigen meinen Bruͤdern g. Quam vone Samario ein quena sārio/ scephan thaz uuazzer ( ‘kam’; ‘Samaria’; ‘Ehefrau’; ‘sofort’; ‘schöpfen’) 7.5. The words in this exercise have outdated spellings. Rewrite them, using current spelling conventions. Then transcribe them phonemically and phonetically. a. b. c. d. e. f. g. h. 7.6. Regional variation in colloquial German. Identify the region in which you would expect to hear the following pronunciation in colloquial German. Then provide a phonetic transcription of each word in Standard German. a. [mɛːnə] b. [leːjə] c. [tsuːx] d. [ʁiːʃən] e. [vɪntsɪk] f. [stʁaɪ ̯t] g. [kineːzɪn] h. [tsaɪ ̯tʊŋk] i. [laʊ̯fsteːç] j. [vɛʃpə] 7.7. German dialects. Below are sets of words from a dialect in one of the major dialect areas of German. Identify the dialect area of each set as Low German, Central German, West Upper German, or East Upper German, providing evidence in support of your conclusion. Consult table 7.4 for the features to search for in the words in these sets. When orthographic symbols are used, long vowels are indicated by double letters (, , , and so on). a. [ɔpfə] ‘Apfel’, [viɐ̯s] ‘Füße’, [ guɐ̯d] ‘gut’, [kxɔts] ‘Katze’, [liɐ̯] ‘lieb’, [mɔx] ‘mache!’, [maɪ ̯] ‘mein’, [noi ̯] ‘neu’, [bvɛfɐ] ‘Pfeffer’

168  Historical Developments and Regional Variation b.  ‘Bruder’, ‘Fuß’, ‘lieb’, ‘Haus’, ‘neu’, ‘Pfeife’, ‘Zug’, ‘zwei’ c.  ‘Dorf ’, ‘Haus’, ‘Häuser’, ‘ich’, ‘Kind’, ‘lassen’, ‘machen’, ‘Pferd’, ‘wissen’, ‘Zeit’ d. [abl ̩beːmçɛr] ‘Apfelbäumchen’, [hai ̯sjə] ‘Häuschen’, [kalp] ‘Kalb’, [mai ̯] ‘mein’, [miliç] ‘Milch’, [peːl] ‘Pfahl’, [peniŋ] ‘Pfennig’, [ ʃdroːs] ‘Straße’, [dswondsiç] ‘zwanzig’ e.  ‘gleich’, ‘gut’, ‘helfen’, ‘Haus’, ‘Kat­ ­ze’, ‘können’, ‘Kuchen’, ‘Müdigkeit’, ‘mein’, ‘neun’, ‘suchen’, ‘zwanzig’ f.  ‘brechen’, ‘das’, ‘laufen’, ‘mein’, ‘Haus’, ‘schmeißen’, ‘zwei’, ‘zwölf ’, ‘was’ g.  ‘Apfel’, ‘aus’, ‘Bruder’, ‘das’, ‘Dorf ’, ‘gut’, ‘Haus’, ‘ich’, ‘lieb’, ‘machen’, ‘mein’, ‘Pfund’, ‘Zeit’

Chapter 8

Phonetics in the Real World

8.1. Introduction The goal of this chapter is to provide you with tools and exercises that will help you improve your pronunciation of German. You will learn how to use the spelling of German to determine how words in German should be pronounced. You will put this knowledge to use by reading phonetic transcriptions and by producing phonetic transcriptions of German words and phrases. Listening and speaking exercises in this chapter will focus on the “new” consonants and vowels in German, that is, those sounds that have no counterparts in North American English. 8.2 . Orthography and the IPA: Letters and Sounds As you learned in chapter 1, German orthography is relatively transparent. The way words are spelled in German typically tells us how to pronounce them. However, there is not a perfect one-to-one correspondence between the orthographic symbols used to spell words and the sounds used to pronounce these words. A letter can represent more than one sound. For example, can represent both [b], as in ‘but’, and [p], as in ‘thief ’. One sound can be represented by different letters. The sound [ɛ], for example, can be represented by both and , as in ‘bed’ and ‘older’. In addition, two letters can represent one sound (the letters represent [x] in ‘book’) and one letter can represent two sounds ( represents the sounds [ts] in ‘time’). Nonetheless, there are some generalizations that can be made about German orthography that will help you determine which sounds you should be aiming for when you pronounce German words. 8.2 .1. Consonants This discussion of German orthography and consonant sounds follows the order of the discussion of German consonants in chapter 1. We start by treating the letters used to represent single sounds, beginning with plosives, the consonants at the top of the chart in table 1.1, and ending with the letter , which represents the lateral approximant, the consonant at the bottom of the chart. We then talk about the letters used to represent important combinations of consonants like affricates. 169

170  Phonetics in the Real World In our discussion of consonants we will need to talk about syllables. Although you probably know enough about syllables to understand what we will say in this chapter, we want to make sure that you do. If you keep in mind the following points, you should be able to follow our discussion. The sounds in a word are grouped together into larger units that we call syllables. Words like ‘around’, ‘there’, ‘path’, and ‘pointed’ all consist of one syllable; words like ‘but’, ‘berry’, ‘to strive’, and ‘(you) find’ are all made up of two syllables; words like ‘to receive’, ‘apartments’, and ‘certainty’ all have three syllables— and so on. You should have no trouble identifying the number of syllables in a word (it can help to clap your hands when you say a word slowly). The different parts of a syllable are important, and there are special terms for these parts (see chapters 4 and 11 for details). All you need to know for the discussion here, however, is whether a consonant is at the beginning of a syllable or at the end. In a word like ‘to live’ [leːbən], for example, you need to know that the sound is at the beginning of the second syllable. In a word like ‘lively’ [leːphaft], on the other hand, you should recognize that the is at the end of the first syllable. The position in a syllable is particularly important for plosives, the first type of consonant we will discuss. When the letters

, , and are used alone or doubled, they are pronounced [p], [t], and [k], respectively. When , , and appear at the end of a word or a syllable, they are also pronounced as [p], [t], and [k], the voiceless plosives. (1)

Voiceless plosives [p]

‘breakdown’ [panə], ‘folder’ [mapə],  ‘praise’ [loːp], ‘to depart’ [ʔapfaːʁən], ‘have’ [haːpt] [t] ‘tea’ [teː], ‘please’ [bɪtə], ‘song’ [liːt],   ‘friendly’ [fʁɔɪ ̯ntlɪç], ‘in the evening’ [ʔaːbənts], ‘city’ [ ʃtat], ‘topic’ [teːma] [k] ‘comb’ [kam], ‘ceiling’ [dɛkə], ‘day’  [taːk], ‘evidence’ [tsɔɪ ̯knɪs], ‘says’ [zaːkt]

If there are other consonants between , , and and the end of a word or syllable, these letters are still pronounced as [p], [t], and [k], as the words , , and in (1) demonstrate. There are a few additional ways in which [t] is represented in German that you should be aware of. These have therefore been included in (1). The combinations and both represent [t], as the words and show. Although the glottal stop, [ʔ], is a sound of German, it is not represented in the German spelling system. We say more about this sound in chapters 3 and 10.

Phonetics in the Real World  171 As long as they do not occur at the end of a word or syllable, the letters , , and , alone or doubled, represent the voiced plosives, [b], [d], and [ g]. (2)

Voiced plosives [b] [d] [ g]

‘ball’ [bal], ‘liver’ [leːbɐ], ‘ebb tide’ [ʔɛbə] ‘then’ [dan], ‘tired’ [myːdə], ‘ram’ [vɪdɐ] ‘goose’ [ gans], ‘violin’ [ gaɪ ̯gə], ‘excavator’  [bagɐ]

Remember that plosives in word-final and syllable-final position are always voiceless. This also holds for fricatives, as we will discuss in more detail below. The devoicing of plosives and fricatives in these positions is a salient feature of German. You have probably noticed that when different endings are added to a word that change the position of , , or in the word, this can result in alternations of voiced and voiceless plosives in different forms of the word. (3)

Alternations in voicing: , , [b]/[p] ‘verbs’ [vɛɐ̯bən] [d]/[t] ‘rounder’ [ʁʊndɐ] [ g]/[k] ‘wars’ [kʁiːgə]

‘verb’ [vɛɐ̯p] ‘round’ [ʁʊnt] ‘war’ [kʁiːk]

It is important to pay attention to the position of , , and and remember that these letters are pronounced [p], [t], and [k] when they appear at the end of a word or syllable. The representation of the nasals is relatively straightforward. The letter represents [m] when used alone or doubled. The letter , alone or doubled, represents [n]—an exception to this being situations in which occurs before a or . When appears before , the combination of letters is typically pronounced as the velar nasal [ŋ]. When appears before a , the is typically pronounced as the velar nasal, and the is pronounced as the voiceless velar stop, [k]. (4) Nasals [m] [n] [ŋ] , (before )

‘mother’ [mʊtɐ], ‘always’ [ʔɪmɐ] ‘nose’ [naːzə], ‘chin’ [kɪn] ‘long’ [laŋ], ‘bench’ [baŋk]

When the letter appears (alone or doubled) or in the sequence before a vowel in the same syllable, it is pronounced as a consonant—either as [r], [ʀ], or [ʁ]. We will use the voiced uvular fricative, [ʁ], in the examples we provide here, since it is the dominant pronunciation of consonantal . (5)

Consonantal [ʁ] ‘speech’ [ʁeːdə], ‘fright’ [ ʃʁɛk], ‘berry’  [beːʁə], ‘barrier’ [ ʃpɛʁə], ‘Rhine’ [ʁaɪ ̯n]

172  Phonetics in the Real World When (alone or doubled) follows a vowel in the same syllable (with the exception of the vowel in the ending ; see below), it has a vocalic pronunciation; it is pronounced as dark schwa, [ɐ̯]. It is marked with the diacritic [  ̯ ] because the vowel preceding it is the more prominent vowel sound in the syllable. We will refer to this sound as nonsyllabic dark schwa. (6)

Nonsyllabic vocalic [ɐ̯] ‘me’ [miːɐ̯], ‘marvelous’ [hɛɐ̯lɪç], ‘there’  [dɔɐ̯t], ‘(you) drove’ [fuːɐ̯st]

When appears in the ending at the end of a syllable, the entire ending is pronounced as dark schwa, [ɐ]. You can find this ending in nouns like ‘reader’. It is also a plural ending, as in the plural form for the noun ‘man’, namely, ‘men’. It is the ending used to create the comparative form of adjectives, ‘smaller’, as well as an adjective ending, as in ‘good’. Some verb stems have this ending, which is pronounced as dark schwa when followed by one or more consonants. (A verb stem is the infinitive without the or . The stem of the verb ‘to smile’, for example, is ; the stem of ‘to sell’ is .) (7)

Vocalic [ɐ] ‘smoker’ [ʁaʊ̯xɐ], ‘pictures’ [bɪldɐ],  ‘bigger’ [ gʁøːsɐ], ‘big’ [ gʁoːsɐ], ‘to supply’ [liːfɐn]

Although dark schwa is a vowel, we treat it here in our discussion of consonants because it occurs only when the consonant symbol appears. How to pronounce The position of in a syllable determines whether it has a consonantal or vocalic pronunciation. When it occurs before a vowel in the same syllable, it is pronounced as a consonant, as in the word ‘to call’ [ʁuːfən]. When it occurs after a vowel in the same syllable (with the exception of the vowel in the ending ), it is pronounced as nonsyllabic dark schwa, as the word ‘animal’ [tiːɐ̯] shows. When it occurs in the ending (at the end of a word or syllable), the entire ending is pronounced as dark schwa, [ɐ], as we see in the word [kɛltɐ]. When you see a word with the letter , you will need to pay attention to the structure of the word (determine whether it has the ending ) as well as the position of in the syllable. Note: The Duden pronunciation dictionary (Mangold 2005) uses [r] to transcribe consonantal pronunciations of . In addition, Duden transcribes as a consonant when it occurs after short vowels ( ‘form’ [fɔrm]); we transcribe it as a vowel ([fɔɐ̯m]). Both pronunciations are acceptable (Mangold 2005:54).

Phonetics in the Real World  173 It is not as easy to make generalizations about the letters used to represent the fricatives in German as it is for the plosives. For this reason we will start our discussion with the letters used to represent the fricatives produced at the front of the mouth and work our way to those used for fricatives at the back. The letters (alone or doubled) and represent the voiceless labiodental fricative, [f ]. The sequence , which often occurs in words that were borrowed from Greek, also represents the sound [f ]. (8)

[f ]

‘five’ [fʏnf ], ‘suitcase’ [kɔfɐ],  ‘father’ [faːtɐ], ‘active’ [ʔaktiːf ], ‘phrase’ [fʁaːzə]

An important feature to note about is that it represents [f ] when it appears at the beginning of native German words, like the word . Some other words that demonstrate this use of are ‘a lot of ’ [fiːl], ‘bird’ [foːgəl], ‘to forbid’ [fɛɐ̯biːtən] (which has the prefix ), ‘full’ [fɔl], and ‘in front of ’ [foːɐ̯]. The letters and represent the voiced labiodental fricative, [v]. (9)

[v]

‘when’ [van], ‘villa’ [vɪla],  ‘November’ [novɛmbɐ]

As the word shows, typically represents [v] when it occurs at the beginning of words borrowed from another language. Some other examples are words like ‘vase’ [vaːzə] and ‘vowel’ [vokaːl]. When appears word-internally, as in , it typically represents [v]. Whenever occurs at the end of a word or syllable, however, it is pronounced as [f ]. The word in (8) above demonstrates this. Because all fricatives that appear at the end of a word or syllable are voiceless (like all plosives), can only represent [f ] when it occurs in this position. Notice that because can be pronounced as both [v] and [f ], this can result in alternations of [v] and [f ] in different forms of a word. For example, when adjectives or nouns ending in have an ending that begins with a vowel, which causes to become word-internal and syllable-initial, is pronounced as [v]. When these words do not have an ending and is word-final, it is pronounced [f ]. (10) Alternations in voicing: [v]/[f ] ‘more well-behaved’ [bʁaːvɐ] ‘archives’ [ʔaɐ̯çiːvə]

‘well-behaved’ [bʁaːf ] ‘archive’ [ʔaɐ̯çiːf ]

As we will see below, this is a characteristic of that is shared by the symbol . The letters , alone or doubled, and are used to represent the voiceless alveolar fricative, [s].

174  Phonetics in the Real World (11)

[s]

‘the’ [das], ‘to eat’ [ʔɛsən], ‘to pour’  [ giːsən]

Although the symbol is used to represent the sound [s], as in English, it is also used to represent the sound [z], the voiced alveolar fricative. This is the only letter used to represent [z]. (12) [z]



‘to sing’ [zɪŋən], ‘evil’ [bøːzə]

It is important to pay attention to the location of a single in a word because its location will determine how it is pronounced. A single is pronounced as [z] if it occurs at the beginning of a word or a syllable and is followed immediately by a vowel, as the words and in (12) show. If it appears at the end of a word or syllable (with or without consonants between it and the end of the word or syllable), it is pronounced as [s]. The word in (11) and a word like ­ ‘(you) come’ [kɔmst] demonstrate this use of the symbol. Because can be pronounced as both [z] and [s], this can result in alternations of [z] and [s] in different forms of a word, as the following examples demonstrate. (13) Alternations in voicing: [z]/[s] ‘glasses’ [ gleːzɐ] ‘glass’ [ glaːs] ‘frizzier’ [kʁaʊ̯zɐ] ‘frizzy’ [kʁaʊ̯s] ‘to read’ [leːzən] ‘reads’ [liːst] Alternations of voiced and voiceless plosives and fricatives If you look at the alternations presented in (3), (10), and (13), you will notice that the letters , , , , and represent voiced sounds—unless they appear at the end of a word or syllable, in which case they represent voiceless sounds. The in ‘to love’, for example, is pronounced as [b], a voiced bilabial plosive, but in the past participle of this verb, ‘loved’, it is pronounced as [p], a voiceless bilabial plosive. We refer to the alternation as final devoicing. You will need to pay close attention to the position of , , , , and in order to determine how these letters are pronounced.

The three-letter combination is typically used to represent the postalveolar voiceless fricative, [ ʃ ]. The letter is also used to represent this sound when it occurs at the beginning of a word before

and . (14) [ ʃ ] , (before

and )

‘beautiful’ [ ʃøːn],  ‘pointed’ [ ʃpɪts], ‘stone’ [ ʃtaɪ ̯n]

Note that at the beginning of a word before represents [s], not [ ʃ ]: ‘scandal’ [skandaːl].

Phonetics in the Real World  175 The letters and are used to represent the sound [ʒ], which is the voiced counterpart of [ ʃ ]. This is not a common sound in German and occurs only in words borrowed from French and Italian. (15) [ʒ]

‘jelly’ [ʒəleː], ‘disgrace’ [blamaːʒə],   ‘journal’ [ʒʊɐ̯naːl]

The letters and are used to represent [ʒ] when it occurs at the beginning of a word; is used to represent it when it occurs word-internally. The letters are used to represent [ç], the voiceless palatal fricative. The letter in the ending is also used to represent this sound as long as a vowel does not follow it. (16) [ç] , (in )

‘China’ [çiːna], ‘I’ [ʔɪç],  ‘milk’ [mɪlç], ‘many a’ [manç], ‘roll’ [bʁøːtçən], ‘calm’ [ʁuːɪç]

If a vowel follows the suffix, the is pronounced as expected—as the voiced velar stop [ g]: ‘calmer’ [ʁuːɪgɐ]. The combination is also used to represent [x], the voiceless velar fricative. (17) [x]

‘stream’ [bax], ‘but’ [dɔx], ‘book’ [buːx],   ‘also’ [ʔaʊ̯x]

The position of the letters in a word will tell you how to pronounce them. When they appear after the vowel symbols , , , and , they represent the sound [x]. In all other locations they represent the sound [ç]. Which sound is represented by ? In native German words, the sequence can be used to represent the voiceless palatal fricative, [ç], as well as the voiceless velar fricative, [x]. The simplest way to remember how is pronounced is to learn when it represents the velar fricative, [x], which is when it appears after the vowels , , , and . In all other positions, represents the palatal fricative, [ç]. The following quote from Friedrich Nietzsche (1975) may help you remember when represents [x]: Ein gutes Buch macht auch noch seine Gegner geistreich. ‘A good book makes even its opponents intelligent.’

The last fricative, the glottal fricative, [h], is represented by the letter . (18) [h]



‘harbor’ [haːfən], ‘to keep’ [bəhaltən]

176  Phonetics in the Real World The letter is also used to signal that the preceding vowel is long, as in the word ‘to drive’, and is not pronounced. This use of will be taken up in the following section. The alveolar approximant, [ j], is represented by the symbol . (19) [ j]



‘yes’ [ jaː], ‘kayak’ [kaːjak]

The lateral approximant, [l], is represented by the letter , either alone or doubled. (20) [l]



‘lip’ [lɪpə], ‘to want’ [vɔlən]

So far we have discussed the representation of single consonant sounds. There are also some consonant combinations that we will need to discuss because of the way they are represented orthographically. These consonant combinations include [kv] and [ks] as well as the affricates, [pf ], [ts], [tʃ], and [dʒ]. With the exception of the affricate [pf ], you do not yet have all the information you need to be able to determine their pronunciation from the ways in which they are spelled. (21) Affricates and other consonant combinations [pf ] ‘pound’ [pfʊnt] [ts] , ‘time’ [tsaɪ ̯t], ‘pizza’ [pɪtsa],  (before )  ‘cat’ [katsə], ‘nation’ [natsi ̯oːn] [tʃ] ‘to clap’ [klatʃən] [dʒ] ‘jungle’ [dʒʊŋəl] [kv] ‘source’ [kvɛlə] [ks] ‘witch’ [hɛksə], ‘six’ [zɛks]

One important point to note about and is that they are both single letters that are used to represent a sequence of two consonant sounds. You will need to keep this in mind in particular with , which represents [ts] in German, not [z] as in English. This will be clear when you compare the pronunciation of the words for ‘zoo’ in German and English: German [tsoː]; English [zuː]. The use of to represent [ks] is straightforward if you remember that it is also used in English in the same way, as words like , , and demonstrate. The use of to represent [ks] is a bit trickier because this same combination of letters can also represent [çs] or [xs]. However, if you analyze the structure of the following words, you will see that represents [ks] only when and belong to the stem of a word; not when is part of an ending: ‘to grow’ [vaksən], ‘highest’ [høːçst], ‘(you) laugh’ [laxst]. In a word like , for example, belongs to the adjective stem, and belongs to the superlative ending, , so the combination is not pronounced [ks]. The preceding section, especially the information presented in the boxes, should make you realize how helpful German spelling can be as you figure out how to pronounce a word. You can also refer to table 8.1, which provides a list

Phonetics in the Real World  177 Table 8.1 Sound-Grapheme correspondences for German consonants. Sound

[p] [b] [t] [d] [k] [g] [m] [n] [ŋ] [f] [v] [s] [z] [ʃ] [ʒ] [ç] [x] [ʁ] [h] [j] [l] [pf] [ts] [tʃ] [dʒ] [kv] [ks]

Grapheme





, (before )



, (before

and )

, (in )



, (before )



Examples

, , , , , , , , , , , , , , , , , , , ,

, , , ,

, ,

,

, , ,

,

of the consonant sounds of German and the letters typically used to represent them—together with relevant examples. The table also includes affricates and other important consonant combinations. The glottal stop, [ʔ], is not included in this table because it is not represented in the German spelling system. 8.2 .2. Vowels Like many of the letters used to represent consonants, those used for vowels can represent two or more different sounds. The letter , for example, can be used to represent the long tense vowel [oː] as well as its short lax counterpart [ɔ]. (22) [oː] [ɔ]



‘oven’ [ʔoːfən] ‘open’ [ʔɔfən]

Even though the same letter is used to represent these two different vowels, we know which one to use when we pronounce these words. The consonant symbols following the letter provide that information. Notice that there is one

178  Phonetics in the Real World c­ onsonant, , after the long vowel [oː], but two consonants, , after the short vowel [ɔ]. The general rule of thumb that allows us to use consonant symbols to tell us how to pronounce vowels is the following. (23) Rule of thumb for pronunciation of vowel letters One or no consonant letter in the same morpheme indicates that the preceding vowel is long. Two or more consonant letters in the same morpheme indicate that the preceding vowel is short.

A morpheme is the smallest unit of language that has meaning. The word ‘tables’, for example, is made up of two morphemes, ‘table’ and , a plural suffix, which means something like ‘more than one’. A verb form like ‘spends’ is made up of three morphemes, , , and . Although it can sometimes be difficult to formulate the meaning of a morpheme (a good example of this is the prefix ), it is generally not difficult to identify morphemes, and this is all that is necessary for using the rule of thumb in (23). This rule of thumb accounts for the pronunciation of the vowels in the following words. (24) One or no consonant letter after the vowel [oː] ‘where’ [voː] no consonant [uː] ‘foot’ [fuːs] one consonant [eː] ‘whom’ [veːn] one consonant (25) Two or more consonant letters after the vowel [ɔ] ‘to hope’ [hɔfən] [ʊ] ‘sugar’ [tsʊkɐ] [ɛ] ‘narrow’ [ʔɛŋ]

When we count consonant symbols in verbs, we exclude consonants that belong to endings because endings are separate morphemes. All forms of verbs with stems that end in a single consonant letter have vowels that are long. (26) One consonant letter after the vowel in a verb stem [oː] ‘to get’ [hoːlən], ‘(you) get’ [hoːlst], ‘gets’ [hoːlt] [uː] ‘to call’ [ʁuːfən], ‘(you) call’ [ʁuːfst], ‘calls’ [ʁuːft] [eː] ‘to peel’ [ ʃeːlən], ‘(you) peel’ [ ʃeːlst], ‘peels’  [ ʃeːlt]

Verb forms like ‘(you) get’ and ‘gets’, for example, which both have long sounds, show that the consonant letters of endings do not count when determining vowel length.

Phonetics in the Real World  179 We also exclude the letter when we count consonant letters. In the following examples, is not pronounced but can be viewed as a redundant (extra) signal that the preceding vowel is long. (27) as signal of length [eː] ‘ache’ [veː] [aː] ‘saw’ [zaːən] [iː] ‘him’ [ʔiːm]

no consonant no consonant one consonant

Another orthographic convention that is used to indicate vowel length is the doubling of the vowel letters themselves. The double vowels found in German are , , and . (28) Doubled vowels as a signal of length [eː] ‘tea’ [teː] [oː] ‘boat’ [boːt] [aː] ‘pair’ [paːɐ̯]

The final orthographic means of signaling length is the use of following to indicate that the is long. The sequence represents the vowel [iː]. (29) as a signal of length [iː] ‘how’ [viː], ‘to offer’ [biːtən], ‘to squeak’  [kviːtʃən]

We refer to the principle in (23) as a “rule of thumb” because it is not foolproof; it does not work in all cases. In the vast majority of cases, vowel sounds that occur before the letters are short—which is what our rule of thumb predicts. (30) Short vowels before (three consonant letters) [ɪ] ‘to mix’ [mɪʃən] [ɛ] ‘ash-tree’ [ʔɛʃə] [a] ‘bottle’ [flaʃə]

There are exceptions, however. Two exceptions that are worth mentioning are ‘niche’ [niːʃə] and ‘washed’ [vuːʃ]. The vowel sounds that occur before the letters , which are used to represent [ç] and [x], can be either long or short. (31) Long vowels before [uː] ‘to search for’ [zuːxən] [oː] ‘high’ [hoːx] ‘after’ [naːx] [aː]

180  Phonetics in the Real World (32) Short vowels before [ɪ] ‘I’ [ʔɪç] [ɔ] ‘still’ [nɔx] ‘roof ’ [dax] [a]

In the case of , then, you will not be able to use German spelling to determine the length of the preceding vowel. You will have to pay attention to spoken ­German or consult a dictionary to determine whether the vowel is long or short. One additional feature of our rule of thumb is that it cannot be applied reliably to the vowels in prefixes and suffixes. A number of prefixes and suffixes have short vowels but end in a single consonant or no consonant. Some common prefixes and suffixes that do not follow our rule of thumb are the following. (33) Prefixes with short vowels before one or no consonant [a] [ʔap], [ʔan] [ɛ] [ʔɛɐ̯], [fɛɐ̯], [tsɛɐ̯] [ə] [bə], [ gə] [ʊ] [ʔʊm], [ʔʊn] (34) Suffixes with short vowels before single consonants [ɪ] [ɪç], [ɪn], [nɪs] [ə] [çən]

Given the relatively small number of these prefixes and suffixes, it should not take long to learn which sounds their vowel letters represent. Although our rule of thumb is not a reliable predictor of the pronunciation of vowels in prefixes and suffixes, there is one very useful generalization that can be made about the vowel when it occurs at the end of a word or in an ending. As long as does not occur in the ending at the end of a syllable, it will be pronounced as [ə] when it occurs at the end of a word or in an ending. (35) [ə]

‘sun’ [zɔnə], ‘(I) come’ [kɔmə],  ‘opens’ [ʔœfnət], ‘to rest’ [ʁuːən], ‘smaller’ [klaɪ ̯nəʁə]

Rule of thumb for pronunciation of vowel letters One or no consonant letter in the same morpheme indicates that the preceding vowel is long ( ‘to pray’ [beːtən]). Two or more consonant letters in the same morpheme indicate that the preceding vowel is short ( ‘beds’ [bɛtən]). See appendix D for a list of common exceptions to this rule of thumb.

Phonetics in the Real World  181 Recall that the combination at the end of a syllable is pronounced as dark schwa, [ɐ]. This is why in the word , for example, is pronounced as dark schwa: [ ʃøːnɐ]. In the word , in comparison, the appears before a vowel in the same syllable, so it is pronounced as [ʁ], and both letters are pronounced as schwa: [ ʃøːnəʁə]. Table 8.2 provides a list of letters that represent the vowel sounds of German and the sounds that these letters represent. If you use this chart together with the information provided above (the rule of thumb in [23] and in the box; other means of indicating length; the pronunciation of vowels in prefixes and suffixes), you will find that you can determine the pronunciation of the vast majority of common words in German simply by paying attention to how they are spelled. For example, the chart tells us that the vowel can be pronounced either as [øː] or [œ]. In the word ‘to disturb’, we know that it is pronounced as the long tense vowel [øː] because only one consonant letter follows it. Because the in is in an ending, it will be pronounced as schwa, [ə]. Even though there are two consonant letters after the in the word , and , we know that it represents the long tense vowel [øː] because only one consonant, the , follows it in the verb stem. Table 8.2 Grapheme-Sound correspondences for German monophthongs. Grapheme









Sound

[aː] [a] [eː] [ɛ] [ə] [iː] [ɪ] [oː] [ɔ] [uː] [ʊ] [eː] [ɛ] [øː] [œ] [yː] [ʏ] [yː] [ʏ] [ɐ] [ɐ̯]

Example















182  Phonetics in the Real World Table 8.3 Grapheme-Sound correspondences for German diphthongs. Grapheme



Sound [aɪ ̯] [aʊ̯] [ɔɪ ̯]

Examples

, , ,

,

Diphthongs in German are typically represented by sequences of two vowels and therefore easy to identify by the way they are spelled. The common vowel sequences used to represent the three main diphthongs in German are provided in table 8.3. The only two-vowel sequences that pose particular problems for students learning German are and . The first one, , represents the monophthong [iː], as in , which means ‘thief ’. Notice that both and have essentially the same vowel, [iː], and both are spelled with the sequence . The second two-vowel sequence, , represents the diphthong [aɪ ̯], as in ‘rich’. If you remember that as well as can represent this diphthong (, ), this could help you remember how is pronounced. 8.2.3. Advanced Topics In this section we will provide you with some additional information about German orthography and pronunciation. These details go beyond the basics, but they may answer questions that could arise as you become more proficient in the language. You now know that the in the ending is pronounced as [ç] as long as it is not followed by a vowel. It is pronounced as [ç] in ‘curly’ [lɔkɪç], for example, but as [ g] in ‘curlier’ [lɔkɪgɐ]. There is one situation that is an exception to this: If there is another [ç] sound following in a word, the is pronounced as [k]. In the word ‘king’, for example, the is pronounced as [ç]: [køːnɪç]. In the word ‘royal’, however, it cannot be pronounced as [ç] because the suffix contains this sound, so is pronounced as [k] (because of final devoicing): [køːnɪklɪç]. When we discussed the nasals, we said that the combination represents the sound [ŋ] and that before also represents [ŋ]. This is correct as long as the and (or ) both belong to the same morpheme. If not, the is pronounced as [n]. In a word like ‘lung’, both and belong to the same morpheme, so the combination is pronounced [ŋ]: [lʊŋə]. In a word like ‘unhappily’, on the other hand, the belongs to the prefix and the belongs to the root, a different morpheme, so the is pronounced as [n]: [ʔʊngɛɐ̯n]. For the same reason, the in ‘unwise’ is pronounced [n], not [ŋ]: [ʔʊnkluːk].

Phonetics in the Real World  183 You may have already noticed that the combination is used to represent more than just the sounds [ç] and [x]. As the following examples show, these letters are also used to represent [k], [ ʃ ], and [tʃ]. (36) [k] [ ʃ ] [tʃ]

‘chorus’ [koːɐ̯], ‘character’ [kaʁaktɐ] ‘boss’ [ ʃɛf ], ‘charming’ [ ʃaɐ̯mant] ‘to check’ [tʃɛkən], ‘to charter’ [tʃaɐ̯tɐn]

You will recognize these words through your knowledge of English, and even though some are originally borrowings from French, you can often use the En­ glish pronunciation of these (or related) words to help you determine how to pronounce the sounds in them that are represented by . For example, the in English is pronounced with a [k], as is the in German . You have no doubt already encountered words in German with the affricate [dʒ] that are not spelled with . This affricate is only found in loanwords, and the ones you will encounter in German will often be familiar to you through your knowledge of English. As the following examples show, and are used to represent this affricate—in addition to . (37) [dʒ]

‘jeans’ [dʒiːns], ‘to jog’ [dʒɔgən] ‘manager’ [mɛnɪdʒɐ], ‘gin’ [dʒɪn]

Again, when you encounter words like these, you can rely on your knowledge of English to help you determine how the consonant letters in these words are pronounced. 8.3. Reading the IPA The following exercises will give you practice in reading the IPA. Commas, semicolons, periods, and spaces between words have been added for ease of reading in those exercises that involve stretches of speech longer than a single word. 8.1. Read the following phonetic transcriptions aloud. Then transcribe the words into standard German orthography. a. [bʁɪŋt] b. [haɪ ̯sɐ] c. [tsuːk] d. [vɛtɐ] e. [bʁaːfstən] f. [ʁuːɪç] g. [haːzə] h. [zaːən] i. [kœntən] j. [liːploːs]

184  Phonetics in the Real World k. [tsaːnflaɪ ̯ʃ] l. [ʔɛltɐ] m. [kvɪtʊŋ] n. [dʊɐ̯çʃaʊ̯ən] o. [faɪ ̯nt] p. [zɔmɐʔaːbənt] q. [bəhɪndɐt] r. [fɛɐ̯liːʁən] s. [mʏstə] t. [zɪçɐ] 8.2. Read the following phonetic transcriptions of German sentences aloud. Then transcribe the sentences into standard German orthography. a. [ʔɪç haːbə maɪ ̯nə tœçtɐ bəzuːxt] b. [diː ʔaʊ̯tobaːn vʊɐ̯də ʔɪn baɪ ̯dən ʁɪçtʊŋən totaːl gəʃpɛɐ̯t] c. [diː tsaːl deːɐ̯ ʔʊnfɛlə ʔaʊ̯f deːn bɛɐ̯liːnɐ gəvɛsɐn ʔɪst ʔɪn deːn lɛtstən jaːʁən  ʃtɛndɪç gəzʊŋkən] d. [ziː vɔlən vɔlf biːɐ̯man høːʁən ʔaːbɐ nuːɐ̯ fʏnfhʊndɐt pasən ʔɪn deːn zaːl] e. [vaʁʊm bɪn ʔɪç ʔɪmɐ zoː myːdə] f. [ziː leːpt ʔɪn deːɐ̯ ʔeːɐ̯stən ʔetaːʒə ʔɪn deːm haʊ̯s ʔan deːɐ̯ ʔɛkə] g. [fʁaŋkfʊɐ̯t ʔɪst diː tɔɪ ̯ɐstə ʃtat dɔɪ ̯tʃlants] h. [ziː ʔɪst vaːnzɪnɪç ʃøːn] i. [baɪ ̯ deːm ʁɔkkɔntsɛɐ̯t vaːʁən ʔam fʁaɪ ̯taːk meːɐ̯ ʔals fʏnftsɪç mɛnʃən fɛɐ̯lɛtst  vɔɐ̯dən] j. [deːn zats bʁaxtə ʔeːɐ̯ nɪçt tsuː ʔɛndə] 8.3. Read the following transcriptions aloud. Each represents the pronunciation of more than one word in German. For each transcription, provide the standard German spelling for as many different words as indicated in parentheses. a. [das] (2) b. [ʁaːt] (2) c. [zaɪ ̯tə] (2) d. [ ʃtiːl] (2) e. [ʁaɪ ̯n] (3) f. [lɔɪ ̯tən] (2) g. [ ʃtat] (2) h. [toːt] (2) i. [bɛlə] (2) j. [maɪ ̯n] (2) 8.4. Read the following excerpt aloud and then transcribe it into standard German orthography. Which word in the second sentence in the second paragraph violates the rule of thumb for the pronunciation of vowel letters and why?

Phonetics in the Real World  185 [diː bʊtɐ ʔɪst das ʔɛɐ̯ʃtaɐ̯tə fɛt deːɐ̯ mɪlç, ʔɛnthɛlt ʔaːbɐ nɔx ʔʊngəfeːɐ̯ fʏnftseːn pʁotsɛnt mɪlç ʔɪn faɪ ̯nstɐ fɛɐ̯taɪ ̯lʊŋ. baɪ ̯m ʃmɛltsən tʁɪt ʔaɪ ̯nə tʁɛnʊŋ ʔaɪ ̯n; diː maːgɐmɪlç zɪŋkt tsuː boːdən, daʁyːbɐ ʃteːt ʔaɪ ̯nə klaːʁə fɛtʃɪçt. lɛst man diːzə ʔɛɐ̯kaltən, zoː ʔɛɐ̯ʃtaɐ̯t das klaːʁə fɛt ʔʊnt hat deːn naːmən ʃmɛltsbʊtɐ ʔoːdɐ bʊtɐʃmalts ʔɛɐ̯haltən. diː ʔɪn deːɐ̯ bʊtɐ ʔaɪ ̯ngəʃlɔsənə mɪlç ʔɪst kaɪ ̯n tsuːfɛlɪgɐ bəʃtanttaɪ ̯l, ʔaʊ̯x kaɪ ̯nə fɛɐ̯ʔʊnʁaɪ ̯nɪgʊŋ ʔoːdɐ fɛɐ̯fɛlʃʊŋ, zɔndɐn ʔaɪ ̯n noːtvɛndɪgɐ bəʃtanttaɪ ̯l deːɐ̯ bʊtɐ, deːɐ̯ ʔeːɐ̯st das bʊtɐfɛt tsuː bʊtɐ maxt.] [deːɐ̯ gəʃmak deːɐ̯ bʊtɐ ʔɪst ʔaphɛŋɪç dafɔn, ʔɔp diː bʊtɐ ʔaʊ̯s zyːsəm ʔoːdɐ zau̯əʁəm ʁaːm heːɐ̯gəʃtɛlt ʔɪst, ʔɔp ziː gəzaltsən ʔoːdɐ ʔʊngəzaltsən ʔaʊ̯f deːn tɪʃ kɔmt. deːɐ̯ voːlgəʃmak deːɐ̯ bʊtɐ kan dʊɐ̯ç fɛɐ̯ʃiːdənə ʔaɪ ̯nflʏsə bəʔaɪ ̯ntʁɛçtɪçt veːɐ̯dən. tsuː laŋəs fɛɐ̯vaɪ ̯lən deːɐ̯ mɪlç ʔɪm ʃtal fɛɐ̯laɪ ̯t deːɐ̯ bʊtɐ ʔaɪ ̯nən ʃtalgəʃmak. diː bʊtɐ vɪɐ̯t zeːɐ̯ laɪ ̯çt ʁantsɪç, vɛn ziː ʔɔfən ʔan deːɐ̯ lʊft ʃteːt, ʔʊnt gants bəzɔndɐs vɛn ziː fɔm lɪçt gətʁɔfən vɪɐ̯t.]

8.4. Transcription I: From the Written Word In the section on orthography and the IPA earlier in this chapter, you learned about the relationship between letters and sounds in German. With this knowledge, use the orthography of German to provide phonetic transcriptions of the sounds, words, and phrases in the following activities. Transcribe consonantal as the voiced uvular fricative, [ʁ]. 8.5. Write the phonetic symbol of all the plosives in the following words. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. p. q. r. s. t.

186  Phonetics in the Real World 8.6. Write the phonetic symbol of all the nasals and plosives in the following words. a. b. c. d. e. f. g. h. i. j. 8.7. Write the phonetic symbol of the sound represented by in the following words. Is it [ʁ] or [ɐ̯]? a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. 8.8. [ç] or [x]? Write the phonetic symbol of the fricative represented by in the following words. a. b. c. d. e. f. g. h. i.

Phonetics in the Real World  187 j. k. l. m. n. o. p. q. r. 8.9. Write the phonetic symbol of the fricatives in the following words. Include the consonantal pronunciation of , which is the fricative [ʁ]. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. p. 8.10. [ g], [k], or [ç]? Write the phonetic symbol for the pronunciation of in the following words. You will need to keep in mind the details about the pronunciation of when it appears in the ending . a. b. c. d. e. f. g. h. i. j.

188  Phonetics in the Real World k. l. m. n. o. p. 8.11. Write the phonetic symbol of all the consonant sounds in the following words. a. b. c. d. e. f. g. h. i. j. k. l. 8.12. [ə] or [ɐ]? Write the phonetic symbol for the vowel sounds in the underlined portions of the following words. Remember that [ɐ] is the pronunciation of at the end of a syllable (with or without consonants between and the end of the syllable). a. b. c. d. e. f. g. h. i. j. k. l. m. n. o.

Phonetics in the Real World  189 8.13. Write the phonetic symbol for the vowels in the following words. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. p. 8.14. Prefixes and suffixes. Write the phonetic symbol for the vowels in the following words. a. b. c. d. e. f. g. h. i. j. 8.15. Long vowels or diphthongs? Write the phonetic symbols for all the sequences of two vowels in the following words. a. b. c. d. e. f. g. h.

190  Phonetics in the Real World i. j. k. l. m. n. o. 8.16. Number of letters versus number of sounds. Provide a phonetic transcription for the following words. Keep in mind that there is not necessarily a one-to-one correspondence between the number of letters and number of different sounds. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. p. q. r. s. t. 8.17. [ʁ], [ɐ], or [ɐ̯]? Transcribe the following words phonetically. Remember that [ʁ] is the pronunciation of before a vowel in the same syllable; [ɐ̯] is the pronunciation of after a vowel in the same syllable; and [ɐ] is the pronunciation of at the end of a syllable (with or without following consonants in the syllable). a. b. c. d. e.

Phonetics in the Real World  191 f. g. h. i. j. k. l. m. n. o. p. 8.18. Transcribe all the words in table 8.1. 8.19. Transcribe all the words in tables 8.2 and 8.3. 8.20. Transcribe the following phrases. For now, you should transcribe a glottal stop, [ʔ], before all word-initial vowels. For example, you should transcribe a word like ‘often’ as [ʔɔft]. See chapters 3 and 10 for more information on the glottal stop. a. weder Fisch noch Fleisch sein b. kein heuriger Hase sein c. aufs Land fahren d. von Luft und Liebe leben e. bittere Not leiden f. ein frühes Grab finden g. lieber Wein als Bier h. in dunkler Vergangenheit i. seinen Hut nehmen müssen j. ein dickes Fell haben k. tanzen und springen l. in Rauch aufgehen m. nichts als Blödsinn reden n. Jubel, Trubel, Heiterkeit o. Tod auf dem Schlachtfeld p. summende Bienen 8.5. Pronunciation Similarities and Differences In the sections that follow, we will discuss the sounds of German as they fit into three groups: those that are roughly the same in North American English and German, those that are different, and those that are typically perceived and treated by language learners as similar in the two languages. Research has shown

192  Phonetics in the Real World that segments that are the same in two languages are not problematic for learners, and those that are different can also be learned. It is those sounds that are similar across two languages that have been shown to pose the greatest difficulty for learners (Flege et al. 2003). The pronunciation and perceptual practice in this chapter will focus on the “new” German segments (those that are different from North American English segments). Consonants that are very much the same in German and English and can therefore be produced essentially alike include the following: [p b t d k g ʔ m n ŋ f v s z ʃ ʒ h j]. This means that in most instances, producing consonantal segments in German as you would in English will work well for you. It is important to remember, however, that these are segments, not letters of the alphabet. This is especially true of obstruents (plosives and fricatives) that occur at the ends of syllables. This aspect of German pronunciation is taken up in greater detail in chapters that follow. For now, though, it is important to remember that the letters at the ends of words and syllables are pronounced as voiceless sounds: [p t k f ]. Although these are not new sounds, this is a new pattern for native speakers of English. This means that German words that end with letters that are typically used to represent voiced obstruents, ‘foliage’, ‘envy’, ‘lay’, and ‘naïve’, are all pronounced with voiceless obstruents: [laʊ̯p], [naɪ ̯t], [leːk], and [naiːf ]. It is also important to remember that the symbol represents [z] in syllable-initial position before a vowel, but [s] in syllable-final position: ‘circles’ is pronounced [kʁaɪ ̯zə], whereas ‘circle’ is pronounced [kʁaɪ ̯s]. The German consonants [ç] and [x] can be considered new segments in that they do not occur in English. The production of these segments and the devoicing of syllable-final obstruents are taken up below. There is one vowel that is nearly identical in its production in German and English: [iː]. In some environments, though, there are slight differences. The new German vowels for native speakers of English include the front, rounded vowels, [ yː ʏ øː œ], and dark schwa, [ɐ]. The production of these segments is taken up below. Unlike the consonants, most vowels of German and English differ at least somewhat from one another. While many of the segments share a common IPA symbol, their production differs significantly across the two languages. In particular, German vowels tend to make greater use of articulatory space than do English vowels. Specifics about the production of the similar vowels are taken up in chapter 10. 8.6. Pronunciation Practice I As we noted in chapter 1, articulatory phonetics is concerned with the movement of the vocal organs to produce speech. Knowing how the sounds of German are produced is a first step to producing them correctly, and practicing may help you as a learner to become more accurate in your production. Before we begin the

Phonetics in the Real World  193 practice exercises, we would like to make a few important points. First of all, it is quite unlikely that you as an adult learner of German will accurately produce all of the segments like a native speaker. In fact, there are some sounds—especially —that stand out in the German speech of native speakers of North American English. We do not say this to discourage learners from working on their pronunciation. Instead, we hope to encourage you to have reasonable expectations and to focus instead on the second point that we would like to raise: comprehensible second language speech is a realistic goal (Derwing and Munro 2009). That is, being understood by other speakers of German is possible, even when your speech has a detectable foreign accent. Research has shown that pronunciation practice can make second language speech more comprehensible, especially if pronunciation training proceeds through a series of steps. It is for this reason that the following pronunciation activities are included throughout the book and appear in the order presented here: a brief description of the production of sounds, isolated practice (using words), and contextualized practice (using sentences, paragraphs, and dialogues). In the practice sections that follow, we will focus on those German segments that are new for speakers of North American English. This means that we will work with the consonants [ç], [x], and syllable-final devoiced obstruents. In terms of vowels, we will focus on the front, rounded vowels, [ yː ʏ øː œ], and dark schwa, [ɐ]. We will begin with a brief overview of each of the phones and a discussion of how they differ from North American English segments. The activities below focus on both perception and production, because it has been shown that a combination of the two can be helpful for learners. 8.6 .1. New German Consonants It is a realistic goal for English-speaking learners of German to acquire the new German phones [ç] and [x] and to learn to devoice obstruents at the ends of syllables. There are two approaches that can be helpful in producing the palatal fricative, [ç]. One approach is to make use of the in English words like , , and . Many speakers of English pronounce the in these words with a sound that is very much like [ç]. Here is how you can use this sound to help you say [ç]: First isolate this sound by starting to say the word [hjuː] and then just prolonging the . If, while pronouncing this , you narrow the space between the front of the tongue and the palate enough so that friction is produced, you will be saying [ç]. To be able to pronounce [ç] after front vowels (one of the environments it which it appears), start by putting the vowel [ɪ] (which occurs in a word like ‘is’) in front of the word . You will first be saying something like [ɪhjuː]. Then say this again, but change the pronunciation of so that you are creating more friction when you say it (by narrowing

194  Phonetics in the Real World the space between the front of your tongue and your palate), which will produce [ɪçjuː]. Then say this word again, but stop after [ç], and you will have said [ɪç], the German word ‘I’. The sequence to repeat is the following: [ɪhjuː], [ɪçjuː], [ɪç]. A second approach to pronouncing [ç] is to work with [ j], the sound that occurs at the beginning of a word like ‘yes’. This sound is palatal, like [ç]. It differs from [ç], however, in that it is voiced. It is also an approximant, not a fricative, so there is no friction when it is produced. If you whisper the word , you will produce a voiceless . If you whisper again, prolong the , and then also narrow the space between the front of your tongue and the palate enough to produce friction, you will be saying [ç]. German [ç] appears at the beginnings of words ( ‘chemistry’ [çemiː]) and at the beginning of the diminutive suffix ( ‘girl’ [meːtçən]), after the front vowels ( ‘light’ [lɪçt]), and after consonants in the same syllable ( ‘which’ [vɛlç]). It is also the pronunciation of in the ending as long as it does not occur before a vowel ( ‘little’ [veːnɪç]). The velar fricative, [x], is produced in the back of the mouth by raising the back of the tongue toward the velum, so it makes physiological sense that it appears after the nonfront vowels. It appears after the central vowels ( ‘broke’ [bʁaːx], ‘subject’ [fax]) and the back vowels ( ‘hole’ [lɔx], ‘scarf ’ [tuːx]). To figure out where to put your tongue to produce [x], you can use the [k] sound after the central and back vowels as a guide. For example, say the word ‘engine’ [lɔk] and feel where the back of your tongue touches the velum. This is where you will want it when you say the [x] in ‘hole’ [lɔx]. Do not allow the back of the tongue to touch the velum, however. Leave some space so that the air can escape with audible friction. You will then be producing the velar fricative, [x]. It is common for native speakers of English to produce [ç] and [x] as [k] or [ ʃ ]. If you find yourself producing as [k] (for example, saying ‘I’ [ʔɪç] as [ʔɪk] or ‘still’ [nɔx] as [nɔk]), you are producing it as a plosive instead of a fricative. Therefore, instead of stopping the air as you produce the sound, allow it to flow. If [ ʃ ] is what you produce (for example, if you say ‘I’ [ʔɪç] as [ʔɪʃ ]), then you are producing the segment too far forward in your mouth and you are probably also rounding your lips. Try repeating the syllable [ jɪ] twice and then follow it with [ jɪç]. This should give you a feel for the proper place of articulation, since [ j] is palatal, like [ç]. Although the production of voiceless obstruents at the ends of syllables and words is common in English (, , ), voiced obstruents are also allowed (, , ). The difference in Modern Standard German is that only voiceless obstruents are produced at the ends of syllables. In a word like ‘songs’ [liːdɐ], for example, the is pronounced as a voiced alveolar

Phonetics in the Real World  195 plosive, [d], but in the singular form, ‘song’ [liːt], it is pronounced as the voiceless alveolar plosive, [t], because it occurs at the end of a syllable. If you find yourself producing voiced obstruents at the ends of syllables, we recommend that you spend time reading aloud in German. Before you begin reading you may wish to underline the obstruents that you will produce as voiceless sounds. A number of the exercises below focus specifically on the devoicing of obstruents in syllable-final position. 8.6 .1.1. Perceptual Practice 1 8.21. Listen to the following words and write which phone ([ç], [x], [ ʃ ], [k]) you hear for each of the underlined segments. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. 8.22. Listen to the following words and fill in the appropriate phone in each. Example: Armban[ t ]uhr a. Freita[ ] b. Urlau[ ]er c. Klei[ ]er d. Lei[ ] e. Verbän[ ]e f. Fie[ ]er g. Gel[ ] h. run[ ]lich i. Stran[ ] j. Hal[ ] k. tä[ ]lich

196  Phonetics in the Real World l. schrei[ ]en m. Para[ ]ies n. bra[ ] o. Far[ ]e 8.23. Listen to the sentences and decide whether the underlined segment is voiced or voiceless. a. Die Frau hat drei Kinder. b. Wenn es kalt ist, trage ich ein Halstuch. c. Wir schwimmen gern im Schwimmbad. d. Die nativen Olivenöle schmecken am besten. e. Ist Kalbfleisch sehr teuer? f. Meine Mutter ist Marburgerin. g. Ärmellose T-Shirts sind in der Schule nicht erlaubt. h. Im Winter trage ich immer Handschuhe. i. Die braven Kinder hören gut zu. j. Sie trägt einen Korb, wenn sie einkaufen geht. 8.24. Listen to the poem that follows. Underline all of the examples of final devoicing that you hear. Do not include the in the word ‘and’. Der Werwolf Ein Werwolf eines Nachts entwich von Weib und Kind und sich begab an eines Dorfschullehrers Grab und bat ihn: Bitte, beuge mich! Der Dorfschulmeister stieg hinauf auf seines Blechschilds Messingknauf und sprach zum Wolf, der seine Pfoten geduldig kreuzte vor dem Toten: Der Werwolf‹ – sprach der gute Mann, ›des Weswolfs, Genitiv sodann, ›dem Wemwolf, Dativ, wie man’s nennt, ›den Wenwolf, – damit hat’s ein End’.‹ Dem Werwolf schmeichelten die Fälle, er rollte seine Augenbälle. Indessen, bat er, füge doch zur Einzahl auch die Mehrzahl noch! Der Dorfschulmeister aber mußte gestehn, daß er von ihr nichts wußte. Zwar Wölfe gäb’s in großer Schar, doch ›Wer‹ gäb’s nur im Singular.

Phonetics in the Real World  197 Der Wolf erhob sich tränenblind – er hatte ja doch Weib und Kind!! Doch da er kein Gelehrter eben, so schied er dankend und ergeben. Christian Morgenstern (1917)

8.25. Listen to the following poem. Above each example of , indicate whether you have heard [x] or [ç]. (Exclude the in , which is pronounced [k].) Bim, Bam, Bum Ein Glockenton fliegt durch die Nacht, als hätt er Vogelflügel; er fliegt in römischer Kirchentracht wohl über Tal und Hügel. Er sucht die Glockentönin BIM, die ihm vorausgeflogen; d.h., die Sache ist sehr schlimm, sie hat ihn nämlich betrogen. “O komm,” so ruft er, “komm, dein BAM erwartet dich voll Schmerzen. Komm wieder, BIM, geliebtes Lamm, dein BAM liebt dich von Herzen!” Doch BIM, daß ihrs nur alle wißt, hat sich dem BUM ergeben; der ist zwar auch ein guter Christ, allein das ist es eben. Der BAM fliegt weiter durch die Nacht wohl über Wald und Lichtung. Doch, ach, er fliegt umsonst! Das macht, er fliegt in falscher Richtung. Christian Morgenstern (1917)

8.26. Listen to the text that follows from Spiegel Online (2012). Underline all of the examples of and indicate whether you have heard [x] or [ç]. Studentische Selbsttäuschung: Juhu, ich langweile mich Das Referatsthema ist öde, das Studium langweilig, der Job ätzend. Doch wir verteidigen alles als spannend, interessant, herausfordernd, vor allem vor uns selbst. ­Stefanie Unsleber, 24, spürt solchen Uni-Alltagsphänomenen nach – und erklärt, wie wir uns selbst austricksen.

198  Phonetics in the Real World Mein erstes Studium habe ich abgebrochen. Islamwissenschaft und Philosophie, in der Universität war das langweiliger als in den Geschichten von “Sophies Welt” und den Infobänden zu islamischem Fundamentalismus. Ich brauchte ein neues Fach. Die Wahl fiel auf Politikwissenschaft. “Ist es das jetzt wirklich?”, fragten meine Eltern. “Bist du dir sicher?”, meine Freunde. “Ja”, sagte ich. Der Studienplan klang gut: Internationale Beziehungen, Vergleich politischer Systeme, Afrika, Asien, Krieg, Krisen und dazu der theoretische Unterbau. Es war schrecklich langweilig. Aber das erzählte ich keinem, auch nicht mir selbst. Stattdessen hörte ich mich über die Interdependenztheorie plaudern. Schilderte, wie interessant die Parteienlandschaft Thailands ist. Versuchte, meine Freunde für das Rechtsstaatsprinzip zu begeistern. Wahrscheinlich habe ich sie ziemlich genervt, mich selbst ja auch. Aber Menschen wie ich begegnen mir ständig. Da ist die Freundin, die sich nie besonders für Betriebspsychologie begeistern konnte, aber nun ständig davon spricht, seit sie ihre neue Hiwi-Stelle hat. Oder meine ehemalige Mitschülerin, die jetzt als Finanzberaterin arbeitet und mir erklärte, wie interessant sie das deutsche Steuersystem findet. © Spiegel Online 2012

8.6.1.2. Production Practice 1 8.27. Read each German word aloud. Then indicate which phone you have produced for the underlined segments and provide a rationale for your production. Example: [ç] following a front vowel, [ɛ] a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. p. q.

Phonetics in the Real World  199 r. s. t. 8.28. The alternations between [ç] and [x] and voiceless and voiced consonants are often found in different grammatical forms of words. Produce the following pairs of words and write down the consonantal alternations that you produce. Example: - [x] - [ç] a. - b. - c. - d. - e. - f. - g. - h. - i. - j. - 8.29. Read “Der Werwolf ” in exercise 8.24 aloud. You may wish to listen to it a few times before you read it aloud. Pay close attention to final devoicing. 8.30. Read the poem from exercise 8.25. Pay particular attention to your production of [x] and [ç]. 8.31. Read the text from Spiegel Online in exercise 8.26 aloud. Concentrate especially on your production of [x] and [ç]. 8.32. Before you read the paragraphs that follow, underline all occurrences of as well as all examples of final devoicing. Pay close attention to these new sounds as you read the paragraph aloud. Elf Söhne Ich habe elf Söhne. Der Erste ist äußerlich sehr unansehnlich, aber ernsthaft und klug; trotzdem schät­ ­ze ich ihn, wiewohl ich ihn als Kind wie alle andern liebe, nicht sehr hoch ein. Sein Denken scheint mir zu einfach. Er sieht nicht rechts noch links und nicht in die Weite; in seinem kleinen Gedankenkreis läuft er immerfort rundum oder dreht sich vielmehr. Der Zweite ist schön, schlank, wohlgebaut; es entzückt, ihn in Fechterstellung zu sehen. Auch er ist klug, aber überdies welterfahren; er hat viel gesehen, und deshalb scheint selbst die heimische Natur vertrauter mit ihm zu sprechen, als mit den Daheimgebliebenen. Doch ist gewiß dieser Vorzug nicht nur und nicht einmal wesentlich dem Reisen zu verdanken, er gehört vielmehr zu dem Unnachahmlichen dieses Kindes,

200  Phonetics in the Real World das zum Beispiel von jedem anerkannt wird, der etwa seinen vielfach sich überschla­ genden und doch geradezu wild beherrschten Kunstsprung ins Wasser ihm nachmachen will. Bis zum Ende des Sprungbrettes reicht der Mut und die Lust, dort aber statt zu springen, setzt sich plötzlich der Nachahmer und hebt entschuldigend die Arme. – Und trotz dem allen (ich sollte doch eigentlich glückselig sein über ein solches Kind) ist mein Verhältnis zu ihm nicht ungetrübt. Franz Kafka (1919)

8.6.2. New German Vowels As is the case with the consonants discussed above, it is possible for you to produce the vowels in this section—in spite of their status as new—if you make slight adjustments to other vowels. You can produce each of the front, rounded vowels, [ yː ʏ øː œ], by articulating their unrounded counterparts, [iː ɪ eː ɛ], and then rounding your lips. It is important to remember that although the practice stage might involve movement from the unrounded to the rounded vowel, German front, rounded vowels are produced as monophthongs, that is, without movement from the unrounded to the rounded counterpart. Although an umlaut (the diacritic placed over a vowel) may seem insignificant, the fronting associated with these vowels is important because it distinguishes meaning. For example, if the in ‘oven’ [ʔoːfən], which is a mid, back, rounded vowel, is umlauted, it becomes a mid, front, rounded vowel, [øː], which results in a different word form: ‘ovens’ [øːfən]. A similar phenomenon in English is the morpheme-final in words like and , which differ in meaning from their counterparts without this : and . If you ignore the pronunciation information provided by this , you will end up saying a completely different word. Similarly, if you ignore the “two dots” over a vowel in German and do not produce the fronting associated with them, you will end up saying something you did not intend to say. We will look now at the pronunciation of each of the front, rounded vowels in German, beginning with [ yː]. A phone similar to German [ yː] exists in the speech of some speakers of North American English. Some speakers produce the in the word with extreme fronting. Even if you are not one of these speakers, you can still learn to pronounce German [ yː]. All you have to do is to produce German [iː] with lip rounding. If you say [iː] and round your lips while saying it, you will end up saying [ yː]. You can also practice this sound by alternating between the German words ‘to lie in a horizontal position’ [liːgən] and ‘to tell a lie’ [lyːgən]. Many native speakers of North American English have difficulty distinguishing [ yː] from [uː]. For example, they have trouble distinguishing ‘goods’ [ gyːtɐ] and ‘good (nominative masculine)’ [ guːtɐ]. If you find that you have difficulty making this distinction, pay close attention to

Phonetics in the Real World  201 the position of your tongue. Say [uː] and then push your tongue forward in your mouth, keeping your lips rounded. You will then be saying the vowel [ yː]. German [ʏ] is produced by making [ɪ] and then rounding your lips. After you practice alternating between [ɪ] and [ʏ], you can practice these sounds in words by saying ‘box’ [kɪstə] and then ‘coast’ [kʏstə]. You may notice that you have the tendency to produce something that sounds like [kʊstə] when you try to pronounce . It is important to remember that this may cause difficulties for native speakers. If you say , for example, instead of , you are saying ‘became’ rather than ‘would’. It is crucial that you focus on the distinction between [ʏ] and [ʊ]. Work on keeping your tongue forward when you pronounce [ʏ]. To produce German [øː], say [eː] and then round your lips. An example of a pair of words that differ only in these vowels is ‘(I) long for’ [zeːnə] and ‘sons’ [zøːnə]. As with the other front, rounded vowels, native speakers of North American English often ignore the umlaut and produce [zoːnə] for the second word in this pair. If you find that you do this, be sure to pay attention to the position of your tongue, and move it forward as you produce the sound. The final front rounded German vowel, [œ], is produced similarly to German [ɛ], but with rounded lips. A pair of words such as ‘to know’ [kɛnən] and ‘to be able to’ [kœnən] is an example of words whose vowels differ primarily in degree of lip rounding. North American English speakers have the tendency to ignore the fronting that goes along with [œ] and often produce the vowel [ɔ] instead. If you find yourself doing this, concentrate on moving your tongue to the position that it is in when you produce the vowel [ɛ]. There are dialects of German, particularly in the south, in which speakers do not produce the front, rounded vowels. Instead, they just produce the front, unrounded counterparts, [i ɪ e ɛ]. This should serve as evidence that merely producing the back “umlautless” vowels [u ʊ o ɔ] will not suffice to distinguish meaning in your speech. The final new vowel in German is dark schwa, [ɐ], which is syllabic when it corresponds to unstressed syllable-final in words such as ‘better’ [bɛsɐ] and ‘mother’ [mʊtɐ]. It is a lax, central vowel that is produced lower in the mouth than [ə]. The North American English in “uh” is a close approximation of this vowel. Your tongue should remain level in your mouth when you produce this segment. One other variant of this vowel is the “r-coloring” that it represents. That is, when follows a vowel in the same syllable in words like ‘clock’, ‘here’, ‘to recognize’ and ‘listens’, the is not produced in the same way as it is at the beginning of a syllable. Instead, it is pronounced as a nonsyllabic version of dark schwa: [ɐ̯]. Because it is a vowel-like pronunciation of , it is often called vocalic . Native speakers of North American English have the tendency to produce an American English [ɹ] whenever they see . If you notice that you are doing this,

202  Phonetics in the Real World remind yourself that whenever is not at the beginning of a syllable, it will be pronounced as dark schwa, a vowel. Syllabic dark schwa, [ɐ], sounds like English , and nonsyllabic dark schwa, [ɐ̯], is simply a less prominent version of this sound that appears when it follows a vowel in the same syllable. 8.6.2.1. Perceptual Practice 2 Instructions for exercises 8.33 through 8.43: Consider each pair of words and circle the word you hear. 8.33. [ yː] or [iː]? a. b. c. d. e. f. g. h.







8.34. [ yː] or [uː]? a. b. c. d. e. f. g. h.







8.35. [ʏ] or [ɪ]? a. b. c. d. e. f. g. h. 8.36. [ʏ] or [ʊ]? a. b. c.

Phonetics in the Real World  203 d. e. f. g. h.





8.37. [ yː] or [ʏ]? a. b. c. d. e. f. g. h.







8.38. [øː] or [eː]? a. b. c. d. e. f. g. h. i.







8.39. [øː] or [oː]? a. b. c. d. e. f. g. h.







8.40. [œ] or [ɛ]? a. b. c. d. e.





204  Phonetics in the Real World f.

g.

h. 8.41. [œ] or [ɔ]? a. b. c. d. e. f. g. h. 8.42. [øː] or [œ]? a. b. c. d. e. f.











8.43. a. b. c. d. e. f. g. h.

i. j. k. l. m. n.

o.

p. 8.44. Listen to each German word and repeat what you hear. Write down the vowel phone that you have heard and produced: [ yː], [ʏ], [øː], or [œ]. a. [ ] b. [ ]

Phonetics in the Real World  205 c. [ d. [ e. [ f. [ g. [ h. [

] ] ] ] ] ]

8.45. Consider each pair of words and circle the word you hear. a.

b.

c.

d. e. f.

g. h.

i.

j. k.

l. 8.46. Listen to each word and fill in the symbol for the phone that you hear. a. Ku[ ]s b. Ei[ ] c. Lag[ ] d. läng[ ] e. Meist[ ] f. Ja[ ] g. lieb[ ] h. Fahr[ ] 8.47. Choose the word that you hear in each of the following sentences. Example: Ich sehe den Kölner / Kellner. a. Der Mann muss das Problem lesen / lösen. b. Die Lehrerin findet ihre Schule / Schüler toll. c. Die Mädchen sehen ihre Mutter / Mütter nicht. d. Da drüben sind die Besen / Bösen. e. “Söhne / Sonne!” sagte die Frau. f. Schüler sollen im Klassenzimmer nicht liegen / lügen. g. Ein Föhn / Phon ist der Linguistin sehr wichtig. h. Das Kind hat zwei Stöcke / Stücke. i. Mein Onkel hat keine Tochter / Töchter. j. Wir mussten / müssten die Antwort sagen.

206  Phonetics in the Real World k. Die Kiste / Küste ist toll. l. “Nein, ich habe Bruder / Brüder gesagt”, sagte der Junge. 8.48. Listen to the following excerpts from “Das Märchen vom Sandmann,” by Hans Christian Andersen (2006). Provide the IPA symbols for each of the underlined segments. a. In der ganzen Welt versteht niemand so schöne Geschichten zu erzählen wie der alte liebe Sandmann. Gegen Abend, wenn die Kinder noch hübsch artig am Tische oder auf ihrem Stühlchen sitzen, kommt das alte Männchen ganz leise die Treppe herauf, denn es geht auf Socken. Husch, öffnet es die Thüre und streut den Kindern Sandkörnchen in die Augen, so fein, so fein, aber doch immer genug, daß sie nicht länger die Augen aufzuhalten vermögen. Deshalb sind sie auch nicht im stande, ihn zu sehen. Er schlüpft gerade hinter sie, bläst ihnen sanft in den Nacken und dann wird ihnen das Köpfchen gar schwer. O ja, aber es thut ihnen nicht weh, denn der Sandmann meint es mit den Kindern gerade gut. Er verlangt nur, daß sie ruhig sein sollen, und das sind sie am besten, wenn man sie zu Bette bringt. b. Über der Kommode hing ein großes Gemälde in einem reich vergoldeten Rahmen, welches eine herrliche Landschaft darstellte. Als der Sandmann dasselbe mit seiner Zauberspritze benetzt hatte, begannen die Vögel darauf zu singen, die Baumzweige bewegten sich, und die Wolken flogen so natürlich, daß man ihren Schatten über die Landschaft konnte dahinschweben sehen. c. „Es fehlt ihnen aber doch unser Grünkohl!“ sagte die Henne. „Ich brachte einen Sommer mit allen meinen Kücheln auf dem Lande zu. Dort war eine Sandgrube, in der wir umhergehen und scharren konnten. Auch hatten wir Zutritt zu einem Garten mit Grünkohl! O wie grün der war! Ich kann mir nichts Schöneres denken!“

8.49. Listen to the paragraphs that follow, from Immensee, by Theodor W. Storm (1852). Provide the IPA symbols for each of the missing phones. Reinhard hatte hier mit Elisabeths Hilfe ein Haus aus Rasenst[ ]cken aufgef[ ]hrt; darin wollten sie die Somm[ ]abende wohnen; ab[ ] es fehlte noch die Bank. Nun ging er gleich an die Arbeit; Nägel, Hammer und die n[ ]tigen Brett[ ] waren schon bereit. Während dessen ging Elisabeth an dem Wall entlang und sammelte den ringf[ ]r­­ migen Samen de[ ] wilden Malve in ihre Sch[ ]rze; davon wollte sie sich Ketten und Halsbänd[ ] machen; und als Reinhard endlich trotz manches krumm geschlagenen Nagels seine Bank dennoch zustande gebracht hatte und nun wied[ ] in die Sonne hinaustrat, ging sie schon weit davon am and[ ]n Ende der Wiese.

8.6.2.2. Production Practice 2 8.50. Practice pronouncing each pair of words in exercises 8.33 through 8.42. For each exercise, pay attention to the differences in tongue position and lip ­rounding

Phonetics in the Real World  207 between the words in each pair. In exercise 8.33, for example, the only difference between the [ yː] in and the [iː] in is lip rounding. In 8.34, the difference is one of tongue position: the [ yː] in is a front vowel, whereas the [uː] in is a back vowel. 8.51. In German, different forms of a word (the singular and plural form, for example) or different words that are related because they share the same root morpheme (a verb and a noun, for example) often exhibit an alternation between different vowels or between [ʁ] and [ɐ̯]. Pronounce the following pairs of words, paying close attention to the vocalic contrasts we have discussed in this chapter as well as the alternation between [ʁ] and [ɐ̯]. a. b. c. d.

e. f.

g. h. i. j.

k. l. m.

n. o. p.

q. r. 8.52. Read each of the following sentences aloud. Be sure to pay close attention to your production of the new vowels. a. Ich habe “Fliege”, nicht “Flieger”, gesagt. b. Es waren vier Kinder, nicht für Kinder. c. Ein Bruder? Nein, zwei Brüder. d. Zwei Öfen habt ihr? Nein, wir haben nur einen Ofen. e. Ich habe “Kissen”, nicht “küssen”, gesagt. f. Das war keine Mühle, sondern ein Müller. g. Es waren viele Fischer aber wenige Fische. h. Ich habe “helle”, nicht “Hölle”, gesagt. i. Das war keine Biene, sondern eine Bühne. j. Er hat “müsste”, nicht “musste”, gesagt.

208  Phonetics in the Real World 8.53. Read aloud the excerpts from Hans Christian Andersen’s “Das Märchen vom Sandmann” in exercise 8.48. Pay close attention to the production of front, rounded vowels when you read. You may wish to listen to the excerpts again before you read them aloud. 8.54. Read aloud the excerpts from Theodor Storm’s Immensee from exercise 8.49. Pay attention to your production of the front, rounded vowels and dark schwa as you read.

Chapter 9

Acoustic Phonetics in the Real World

9.1. Introduction In this chapter you will learn how to use Praat to analyze the German speech of native speakers, which will prepare you to record and analyze your own German speech in subsequent chapters. One of the goals of the chapter is to familiarize you with the tools that make it possible to analyze the acoustic properties of in­ dividual speech sounds. In chapters that follow, you will have the opportunity to analyze additional aspects of speech like word stress and intonation. The exer­ cises in this chapter and in those that follow are meant to provide you with an in­ troduction to acoustic phonetics, and the ultimate goal is to enable you to analyze pronunciation patterns. Simply making use of acoustic analyses will not improve your pronunciation, and the objective of the exercises is not to match the acous­ tic properties of your speech with those of the speech of native speakers. Instead, we hope that by analyzing your own speech, you will come to realize how it dif­ fers from that of native speakers and figure out what you might be able to do to make it more comprehensible when you interact with others in German. After a brief overview of the technology requirements for recording and an­ alyzing speech, this chapter will focus on phonetic analyses. We will begin by gleaning information from waveforms and spectrograms. We will then work on evaluating specific aspects of vowels and consonants by analyzing speech samples. 9.2. Making Use of Computer Technology Software for speech recording and analysis is readily available. Two popular ­programs include Praat (www.praat.org) and Audacity (http://sourceforge.net/ projects/audacity), both of which are available for free online. Given its wide­ spread use among phoneticians, we make use of Praat in our own analyses and when we guide you through the process of acoustic analysis. When you analyze speech, it is possible to evaluate a number of different as­ pects, for example, certain attributes that provide information about the length of segments or manner or place of articulation. It is important to remember, though, that no two speakers produce speech in precisely the same way. In fact, even if a single individual produced the same utterance twice, there would be

209

210  Acoustic Phonetics in the Real World differences in the productions. The following points should guide you in your analyses: 1. Absolute values are less important than the overall patterning of these val­ ues. For example, the duration of long vowels should be longer than that of short vowels. 2. Specific characteristics of consonants may make it difficult to distinguish among certain consonants, especially those with the same manner of ar­ ticulation. Nonetheless, some of the basic characteristics allow us to distin­ guish consonants that differ in manner of articulation (for example, plo­ sives versus fricatives). 3. We speak much more quickly than most of us realize. Therefore, in order to measure aspects of speech accurately, we need to narrow our focus and measure in units that are much shorter than seconds. These units are called milliseconds. One second is equal to 1,000 milliseconds, and one millisec­ ond is equal to 0.001 seconds. We use milliseconds so that we do not have to rely on very small numbers when we talk about the duration of various units of speech. Measuring aspects of speech in seconds would be like mea­ suring your height in miles. If you keep these points in mind, you should be able to learn a great deal about your own speech and the speech of others by analyzing it acoustically. Perhaps you have seen a phonetics lab. It usually has a soundproof booth of some sort along with rather elaborate recording equipment. This equipment al­ lows phoneticians to make high-quality recordings and to later analyze minute details of the speech stream. For your purposes, though, you will not need much specialized equipment to begin recording and analyzing speech. You will need access to a computer, and since Praat works on both PCs and Macs, you should be able to use any computer that is available to you. It is best to work in a quiet room, both when you record and when you analyze speech samples. For record­ ing, the internal microphone on your computer should work well, but you might get better results if you use an external microphone that you can plug into your computer. It also often helps to wear good headphones that cover your ears when you perform your analyses. A microphone and headphones are especially helpful if there are other people in the room where you will make your recordings and perform your analyses. Once you have your hardware ready, the next step is to download and install the most recent version of Praat on your computer. In this chapter we do not discuss many of the technical details of using Praat. These de­ tails are provided in the Praat guide that is available on the book’s Web site. 9 .3. Speech Analysis I: Waveforms and Spectrograms Once we have a recording of a sound, we can view two types of figures to learn more about how it was produced. The first of these is a waveform, also known

Acoustic Phonetics in the Real World  211 as an oscillogram, which is a figure that provides information about a sound’s frequency and amplitude, both of which we will discuss below. We will begin this section with a discussion of frequency as it relates first to pure tones—steady sounds that repeat regularly. Those that repeat more frequently over a unit of time have a higher frequency (a higher pitch) than those that repeat less fre­ quently. On a waveform, time is represented on the horizontal axis (the x-axis). A waveform of human speech is provided in figure 9.1 (see below). 9.1. Listen to the two pure tones (tone a and tone b). Based on what you hear, how do the two tones differ from one another? Which one do you think has a higher frequency? 9.2. Given what you heard when you listened to the two pure tones, how do you expect their waveforms to differ? Now we should test your hypothesis. Begin by downloading the two tones. Open tone a in Praat. Zoom in to 0.1 seconds (100 mil­ l­iseconds) of speech and notice the regularity with which the waveform repeats. Now do the same thing with tone b. In general terms, how would you describe the differences in frequency of the two tones? 9.3. As you know, the sounds of speech are not usually pure tones. Listen to tone c. It is what phoneticians refer to as a Shepard tone, named after the cognitive scientist Roger Shepard. How does it sound compared to the first two tones you listened to? Zoom into 100 milliseconds (0.1 seconds) of tone c. How does the waveform differ in appearance from tones a and b? Although tone c is more complex than tones a and b, it repeats regularly and therefore has a consistent frequency. Because human speech is manipulated in various ways as it makes its way through the vocal tract, it is not made up of a series of regularly repeating waves. We can observe characteristics of specific classes of speech sounds in a waveform. For example, some sounds are louder than others. There is a correlation between the loudness of a sound and its am­ plitude (the degree of change in air pressure). We can observe the differences in amplitude of various speech segments by comparing the height of waves of these segments along the vertical axis (the y-axis). At this point it is helpful to look at a waveform of human speech like that in figure 9.1. The IPA transcription of the word that is pronounced is aligned with the associated segments in the waveform. When we analyze the waveform in figure 9.1, we can begin by distinguishing between sonorant sounds, like vowels and nasals, and obstruents (plosives and fricatives). The waves of the vowels [uː] and [ə] along with the nasal [n], a so­ norant, show the greatest consistency. At the center of each of these segments, where you see the IPA transcriptions for each sound, we see evidence that the waveforms are relatively stable, both in the amplitude of the waves and the regu­ larity of the repetition. We can contrast these sonorants with the plosives [p] and [k] and the fricative [ ʃ ]. Just above the transcriptions of the plosives in figure 9.1

212  Acoustic Phonetics in the Real World

Figure 9.1 Waveform of German ‘to haunt’.

you can observe a spike in the waveform. This corresponds to the burst of air that is released when plosives are produced. Before the transcriptions of the plosives, you will notice that the waveform is relatively flat. The flatness is due to complete stop of the airflow that is characteristic of plosives, or stop consonants. Finally, an important characteristic of the fricative [ ʃ ] is the relative randomness of the waveform. That is to say, the waves are irregular in height, and they do not repeat in a regular fashion. This is due to the turbulent airflow that escapes from the mouth when fricatives are produced. 9.4. Listen to the speech sample. It is the German word ‘floor’. Begin by transcribing this word and classifying each of the consonant segments according to manner of articulation. Based on what you know about what we can observe when we look closely at waveforms, what would you expect to see for each of the segments when you look at the waveform? Consider the following factors for each sound: a. the presence of waves versus silence b. the regularity of the waves c. the amplitude of the waves Now download the file and open it in Praat. To what extent were your hypotheses confirmed? On a speech spectrogram, time is represented on the horizontal axis and matches up with the time represented on a waveform. If you look at the spec­ trogram in figure 9.2, you can compare a number of characteristics of this

Acoustic Phonetics in the Real World  213

Figure 9.2 Spectrogram of German ‘to haunt’.

s­ pectrogram to what you saw in the waveform in figure 9.1. If you look at [uː], [ə], and [n], you will see dark horizontal bands. These dark bands, called formants, are present in sonorous sounds, that is, vowels and sonorant consonants like nasals and [l]. A few formants have been marked in figure 9.2. The darkness of the bands corresponds to the relative amplitude of a segment (the darker the bands, the greater the amplitude and the louder the sound). If you compare the dark bands in sonorants to the amplitude of the same segments in a corresponding wave­ form, you will see that they relate to one another: the higher the amplitude in the waveform, the darker the band in the spectrogram. The location of formants on the y-axis provides us information about how and where sonorants are produced. We will discuss the relationship between formant frequencies and articulation below. Obstruents involve an obstruction of the speech stream, which can be seen on a spectrogram. As you know, plosives are produced by a closure of the vocal tract, blocking the airflow, and then a release of the closure, accompanied by a burst of air. On a spectrogram, like on a waveform, it is easier to begin by locating the re­ lease burst. You can see this in the dark vertical line just above the IPA characters [p] and [k]. The empty space in the spectrogram just before each of these sounds corresponds to the blockage of the airstream. The irregular disturbance above the fricative [ ʃ ] on the spectrogram is evidence of the turbulent noise of fricatives. 9.5. Open the speech sample and spectrogram in Praat. Label each of the segments in the word, based on what you know about the characteristics of each of the segments. 9.4. Analyzing German Vowels Vowels exhibit variability across speakers, and second language learners often show even more variability in their production of vowels than native speakers do. When we analyze vowels, we have to be sure that we are just measuring the vowel

214  Acoustic Phonetics in the Real World

Figure 9.3 Spectrogram of German ‘to haunt’ with steady states of the vowels outlined.

itself, as opposed to the movement into and out of it. We refer to the most stable part of the vowel—the part that shows the least movement—as the steady state. If you look at the spectrograms of the vowels [uː] and [ə] in figure 9.3, you can deter­ mine where the steady state of the vowel is: it is where we see little movement in amplitude, as evidenced by formants that are relatively straight across (formants are indicated in the spectrogram in figure 9.3 by dots). For your purposes, you can rely on Praat to help you determine where the vowel formants are. You can do this by viewing a sound file, clicking on “Formant” and then on “Show formants.” The steady state portions of vowels are highlighted via boxes in figure 9.3. When we talk about formant values, we usually rely on the average values over the duration of the steady state. Some important things to remember when you analyze vowels include the following: 1. The analyses that you can perform on vowels will be relatively imprecise. Therefore, do not worry if the values you measure differ a bit from those provided. If they vary a great deal, you may wish to measure them again. 2. Do not expect the formant values that you measure in your own speech to overlap with those of native speakers. More important than absolute for­ mant values are the patterns of formant values, which indicate the rela­ tive location of vowels within acoustic space. You should produce high vowels higher than low vowels, front vowels further forward in the mouth than back vowels, and rounded vowels with lip rounding. You will learn more about the relationship between formant values and the production of ­vowels below.

Acoustic Phonetics in the Real World  215 When we refer to the quality of vowels, we are really talking about the height and front-to-back position of the tongue and the presence or absence of lip rounding. Although the correlation is not absolute, we can learn about the height of a vowel by measuring its F1 values. Exercise 9.6 will enable you to determine the nature of the relationship between vowel height and F1 values. 9.6. Open the sound file in Praat. It contains a series of mostly nonsense words produced by two speakers. Transcribe the words that you hear. Now open the spectrogram and record the approximate F1 values of the first vowel in each non­ sense word as it is produced by the male and the female speaker. Be sure that you zoom in enough to be able to measure the steady state of the vowels. Enter these F1 values in table 9.1. (Do not worry about having the “correct” values. Approxi­ mate values are sufficient for comparing these vowels.) Which vowels have the highest F1 values? Describe how the F1 values correspond to vowel height. How do the values you measured from the male’s sample differ from those in the female’s sample? Another important feature of vowels that we can measure acoustically is the extent to which they differ along the front-back continuum. We can determine the relative tongue position in the production of vowels by measuring F2 values. You will determine the relationship between frontness and F2 values in the fol­ lowing exercise. Rounding is the final dimension of vowel quality that we will discuss. F2 val­ ues also provide us insight into the roundness of vowels, the nature of which you will discover when you complete the following exercise. One important thing to remember when you produce German vowels, especially [uː], is that rounded Table 9.1 German vowel formant values. Vowel [aː]

[eː] [øː] [iː] [yː] [uː] [oː]

Formant

F1 F2 F1 F2 F1 F2 F1 F2 F1 F2 F1 F2 F1 F2

Female

Male

Your vowels

216  Acoustic Phonetics in the Real World ­ erman vowels are more rounded than are their English counterparts (Hall G 2003:87–88). Therefore, when you produce German vowels that require rounding, be sure that you consciously round your lips. 9.7. Now measure the F2 values of the vowels as produced by both the male and female in the speech sample from 9.6. Enter these values in table 9.1. What do the values you have recorded tell you about F2 values and the front-to-back dimen­ sion? How do the values you measured from the male’s sample differ from those in the female’s sample? 9.8. Now compare the F2 values of the following vowel pairs: [iː] and [ yː], [eː] and [øː]. How do F2 values provide you with information about lip rounding in these vowel pairs? Based on the analyses you performed above, you should have made the f­ollowing generalizations about the relationship between formant values and vowel articulations: 1. The relationship between vowel height and F1 values is an inverse one: high vowels have lower F1 values than low vowels. 2. Front vowels have higher F2 values than back vowels. 3. Lip rounding results in lower F2 values, which means that rounded vowels generally have lower F2 values than their unrounded counterparts. It is often rather difficult for native speakers of English to perceive and pro­ duce differences in vowel duration. This is because differences in the duration of English vowels do not distinguish one vowel from another. Consider, for example, the English words and . Although it might take us longer to produce the [i] in than the [ɪ] in , the vowels—and the words—would be dif­ ferent even if we took the same amount of time to say each vowel. This is because a crucial difference between these two vowels in English is one of tenseness, not duration. In German, on the other hand, duration is a distinctive feature. The duration of the vowels in ‘scarf ’ and ‘sound’ is the crucial feature that distinguishes these words. In addition, all German vowel pairs that differ in tenseness also differ in duration. That is, when vowels occur in stressed syl­ lables, those that are tense are longer than those that are lax. You will have the opportunity to analyze vowel duration differences in chapter 10. Up to this point we have discussed the acoustics of German monophthongs. As you know, German also has three diphthongs, [aʊ̯], [aɪ ̯] and [ɔɪ ̯]. Diphthongs involve a change in quality of a vowel within one syllable; one vowel transitions into another vowel. The shift from the quality of one vowel to the next is evident in a spectrogram, as you can see in figure 9.4. There are two sets of formant struc­ tures, each corresponding to the component vowels that make up the diphthong. Each of the separate sets of formant structures that make up the diphthong [aɪ ̯] is indicated via a box in figure 9.4.

Acoustic Phonetics in the Real World  217

Figure 9.4 Spectrogram of German ‘strike’ with the two components of the diphthong outlined.

A major difference between tense vowels in German and those in North Ameri­ can English is that German tense vowels like the [eː] in and the [oː] in are truly monophthongs, whereas their North American English coun­ terparts like the [eɪ ̯] in and the [oʊ̯] in tend to be diphthongal. We discuss this issue further in chapter 10. 9.9. Open the spectrogram of the sound file, but do not listen to it. Based on what you know about the acoustic properties of diphthongs and monophthongs, de­ termine which vowel is a diphthong and which is a monophthong. Describe the acoustic evidence that you observe in the spectrogram. Then listen to the words and transcribe them. If your predictions were not correct, look again at the spec­ trograms while you listen to the words. 9.5. Analyzing German Consonants When you analyze German consonants, it is important to remember the basic distinction between obstruents (plosives and fricatives) and sonorants (nasals, approximants, and lateral approximants). Obstruents, because they involve ei­ ther complete blockage, in the case of plosives, or substantial constriction along the vocal tract, in the case of fricatives, exhibit relatively random noise associ­ ated with the constriction in both their waveforms and in their spectrograms. So­ norants are more similar to vowels, so they share some characteristics of vowels: more regularly repeating cycles in their waveforms as well as formant structure that is visible in spectrograms. Perhaps the most difficult part about analyzing sonorants is distinguishing them from vowels in the speech stream. Because they share many of the characteristics of vowels, including voicing and formant struc­ ture, it will be important for you to perform your analyses of these vowel-like seg­

218  Acoustic Phonetics in the Real World ments by zooming in on them in Praat. Exercises in some of the following chapters will focus on the acoustic analysis of sonorants. In this chapter, though, you will practice analyzing obstruents. Although it is not difficult to distinguish plosives and fricatives from one an­ other or from other types of segments, especially on a speech spectrogram, it is somewhat more difficult to distinguish among the members of each category. For that reason, the exercises in this section will focus on the broad categories of plosive and fricative. 9.10. The most important features of plosives include a complete blockage of the airstream and a burst of noise, as noted in the discussion of figures 9.1 and 9.2 above. The closure is most visible when the plosive occurs after a vowel. Open the sound file along with its waveform and spectrogram and locate the closure and release burst for the plosives in each of the words. You should be able to locate these both on the waveform and on the spectrogram. Capture the image of either the waveform or spectrogram for one of the words, and label the closure and release burst. As we mentioned above when we discussed figures 9.1 and 9.2, a salient fea­ ture of fricatives is the turbulent noise that is produced when the vocal tract is constricted. You can see evidence of this noise on both a waveform and a spec­ trogram. You should not expect to be able to differentiate the fricatives from one another when you analyze them, but you should be able to locate fricatives more generally, especially on a spectrogram. 9.11. Listen to the speech sample. What are the fricatives you hear in each? Open the sample in Praat. Take a screenshot of the waveform and spectrogram of each word. Label the fricatives in each word, both in the waveform and in the spectro­ gram, by placing a box around them, as we have done in the presentation of the vowels in figure 9.3. In the exercise that follows, you will focus on the production of [ç] and [x] and on differentiating them from sounds that learners often confuse them with. Here is one example of a situation in which phonetic analyses might be able to help you pinpoint precisely where you are making errors. 9.12. Open the sound file and transcribe the two word pairs you hear. Pay close attention to the second consonant in each of the four words. Now open the wave­ form and spectrogram in Praat. Which acoustic differences do you see that cor­ respond to the manner of articulation of the second consonant in each of the words?

Chapter 10

Phonology in the Real World

1 0 .1. Introduction In previous chapters we have said that German orthography is relatively transparent. That is, we can usually predict how a word will be pronounced when we look at its spelling. There are times, though, when words that look similar are pronounced differently, like those in (1). (1)



a. ‘away’ [vɛk], ‘path’ [veːk] b. ‘searches’ [zuːxt], ‘addiction’ [zʊxt]

We call word pairs like those in (1) homographs, since they share the same graphemes. Homographs show us that the same series of letters is not always pronounced in the same way. There are also times when pairs of words are spelled differently but we pronounce them the same, as in (2). We call these word pairs homophones. (2)

a. ‘city’ [ ʃtat], ‘instead’ [ ʃtat] b. ‘offered’ [boːt], ‘boat’ [boːt] c. ‘chorus’ [koːɐ̯], ‘corps’ [koːɐ̯]

Homophones demonstrate that the relationship between letters and sounds is not always a simple one. In this chapter we will work with a type of transcription that can be viewed as the ideal orthography: phonemic transcription. Our goal is to show you how to transcribe words phonemically by using the information provided by graphemes. A second goal of this chapter is to provide you with strategies for improving your perception and production of consonants and vowels that are similar in German and English. By “similar” we mean those sounds in German that native speakers of English perceive and treat as being similar. Research has shown that these consonants and vowels are especially difficult for second language learners, who are often unaware of subtle differences across the two languages. We focus on the consonants , , and the glottal stop, and the vowels , , , and . The acoustic analysis exercises at the end of the chapter will enable you to analyze and fine-tune your own productions of these segments.

219

220  Phonology in the Real World 1 0.2. Minimal Pairs When we worked with the new vowel sounds in chapter 8, many of the perceptual and production exercises involved pairs of words like ‘box’ and ‘coast’, ‘to read’ and ‘to solve’, and ‘brother’ and ‘brothers’. These minimal pairs (pairs of words with distinct meanings that differ by only one sound in the same position in each word) are important for language learners because they identify the sounds of the language that have the potential to change the meaning of a word. The minimal pair [kɪstə] and [kʏstə], for example, shows that [ɪ] and [ʏ] are distinctive sounds in German. If we want to talk about the coast and attempt to say but use the vowel [ɪ] instead of [ʏ], we will end up saying and talking about a box instead of the coast. We call the distinctive sounds of a language phonemes and write them between slashes, //. On the basis of a minimal pair like and , we can say that German has the phonemes /ɪ/ and /ʏ/. Working with minimal pairs, which contrast the distinctive sounds of a language, can help language learners perceive and produce the important differences among the sounds of that language. If you find that you still have difficulty distinguishing and producing some of the new sounds in German, you can revisit chapter 8 and work with the exercises in that chapter that involve minimal pairs. Exercise 8.35, for example, can help you distinguish [ʏ] and [ɪ]. Exercise 8.38 can help distinguish [øː] and [eː]. The exercises in this section will provide additional practice with the new sounds of German as well as practice with other distinctions, for example, the distinction between [aː] and [a] exemplified by ‘steel’ [ ʃtaːl] and ‘stable’ [ ʃtal]. Each of the following exercises requires you to distinguish between two similar but distinct sounds. Exercise 10.1, for example, requires you to distinguish between [x] and [k]. Both are velar, but [x] is a fricative and [k] is a plosive. These two sounds are distinct in German—they function as phonemes. They are capable of changing the meaning of a word, as the minimal pair ‘night’ [naxt] and ‘naked’ [nakt] demonstrates. Each word you hear in these exercises will be one member of a minimal pair. The first word in exercise 10.1, for example, will either be or . You will need to listen carefully to determine which word you hear. Instructions for exercises 10.1 through 10.7: Fill in the appropriate phone in each of the words you hear. Then transcribe each word into standard German orthography and provide its meaning. 10.1. [x] or [k]? a. [na t] b. [a t]

Phonology in the Real World  221 c. [buː ] d. [lɔ ] e. [pa t] f. [dɔ ] g. [la ] h. [tsʊ t] 10.2. [ç] or [ ʃ ]? a. [kɪɐ̯ ə] b. [lœ ɐ] c. [ʔɛntvɪ ən] d. [ʁasɪ ] e. [tɛpɪ ] f. [fɪ tə] g. [zeːlɪ ] h. [vɪ ən] 10.3. [ yː] or [uː]? a. [ g tɐ] b. [bl tən] c. [z t] d. [ʁ dɐ] e. [v st] f. [bl zə] g. [m ən] h. [d stɐ] 10.4. [ʏ] or [ʊ]? a. [v ɐ̯dən] b. [ts ndɐ] c. [v stən] d. [b ɐ̯gən] e. [dʁ kɐ] f. [d ŋkəl] g. [h tə] h. [ ʃt k] 10.5. [øː] or [oː]? a. [l zən] b. [ gʁ sə] c. [h ə] d. [n tə] e. [l nən]

222  Phonology in the Real World f. [t nən] g. [h lə] h. [b gən] 10.6. [œ] or [ɔ]? a. [bl kə] b. [h kɐ] c. [l kən] d. [b kən] e. [ʁ kə] f. [tʁ pfə] g. [ts lə] h. [f ɐ̯dəʁʊŋ] 10.7. [aː] or [a]? a. [l kən] b. [ʁ tən] c. [z t] d. [ ʃt l] e. [k m] f. [h kən] g. [ ʃ l] h. [f l] The following exercise requires you to determine whether two words form a minimal pair. Recall that two words form a minimal pair if they have distinct meanings and differ by only one sound in the same position in each word. 10.8. Transcribe the following pairs of words phonetically and determine whether they form a minimal pair. If they do form a minimal pair, indicate which two sounds contrast. If they do not form a minimal pair, explain why not. a. , b. , c. , d. , e. , f. , g. , h. , i. , j. ,

Phonology in the Real World  223 1 0 .3. Transcription II: From Grapheme to Phoneme In chapter 8 you learned how to use the way in which German words are spelled to help you pronounce them. For example, you learned that a letter (grapheme) like is pronounced as [b] as long as it does not occur at the end of a word or syllable. You also learned that if occurs at the end of a word or syllable, it is pronounced as [p] (because of the process we call final devoicing). This knowledge of grapheme-sound correspondences helped you transcribe words phonetically: you learned to transcribe an inflected adjective like ‘kind’ with a [b], [liːbə], and the uninflected form, , with a [p], [liːp]. In this section we will show you how to use German orthography to transcribe words phonemically. Phonemic transcriptions contain the phonemes of a ­language—the distinctive sounds—and ignore phonetic information that is predictable. For example, all forms of the adjective , inflected and uninflected, can be transcribed phonemically with the phoneme /b/, even though some forms are pronounced with a [p]. Because we can predict when /b/ is pronounced as [p], we do not need to include this information in the phonemic representation of any form of this word. We can always transcribe the stem, (the adjective without any endings), with a /b/: /liːb-/. (3)

/liːbə/ [liːbə] /liːb/ [liːp] (/b/ is pronounced [p] because of final devoicing)

Notice that the phonemic representation of the [b] and [p] sounds in all forms of the adjective is just like the orthographic representation. All forms of are spelled with a , and we represent all forms phonemically with /b/. The use of to spell all forms of the word follows the phonemic principle: (4)

Phonemic Principle One symbol (letter/grapheme) represents one phoneme (distinctive sound).

As you will see, there are many other examples where German orthography follows the phonemic principle. Although this is not always the case—there is not a perfect one-to-one correspondence between all the letters used to spell German and all the phonemes of the language—you can learn to use the way in which words are spelled to transcribe them phonemically. The phonemic representation of a word can be viewed as the ideal way to spell the word. There is no ambiguity if each symbol represents one and only one phoneme, and once you have the phonemic representation of a word, you can use it to determine the actual pronunciation of the word. For example, you can apply what you know about final devoicing to the phoneme /b/. If /b/ appears at the end of a word or syllable, it will be pronounced as [p], a voiceless bilabial ­plosive;

224  Phonology in the Real World otherwise it will be pronounced as [b]: ‘kind’ /liːb/ [liːp]; ‘kind’ /liːbə/ [liːbə]. When we discuss the grapheme-phoneme correspondences in the following sections, we will include information about the pronunciation of phonemes if there is more than one way to realize a phoneme phonetically. 1 0.3.1. Consonants Table 10.1 provides a list of the main grapheme (letter) and consonant phoneme correspondences in German. To simplify the discussion, we have not included some information in the table that you are already well aware of or should have no difficulty determining on your own. By now you know that double consonant letters are used to show that a preceding vowel is short. For example, in the word ‘hall’, indicates that the sound is short: /halə/ [halə]. The double consonant letters do not stand for two identical consonant sounds. (The only time this would happen would be when the consonants belong to two different parts of a complex word—for example, when a prefix ends with the same consonant sound that the rest of the word begins with: ‘seizure’ /aɪ ̯nnaːmə/ [ʔaɪ ̯nnaːmə]). Double consonant letters that are only used to indicate that a preceding vowel is short are not included in table 10.1. We also exclude for /k/ because this combination is also used to show that the preceding vowel is short. The letter is often silent and contributes no information about pronunciation. For example, the combination represents the sound [t] ( /teːmaː/ [teːma]) and represents the sound [ʁ] ( /ʀaɪ ̯n/ [ʁaɪ ̯n]). Neither of these letter combinations is presented in table 10.1 because you can transcribe both as if the were not there: and = /t/; and = /ʀ/. The combination is also not included in table 10.1. It represents the sound [f ] ( [fʁaːzə]), and you should have no trouble transcribing it as the phoneme /f/. We will now give you some information that you will need to use together with the grapheme-phoneme correspondences in table 10.1 to help you transcribe German phonemically. In some cases there are exceptions to the generalizations in the table. In other cases we will point out how you should work with graphemes that are used to represent more than one phoneme. For example, as table 10.1 indicates, a can represent both /f/ and /v/. We will provide information that will help you determine which phoneme to transcribe when a word contains a . We will begin with the graphemes and phonemes at the top of the table. In most cases, when a word has the letters , , and , you can transcribe these as /b/, /d/, and /g/, respectively. The word ‘dress’, for example, can be transcribed phonemically as /klaɪ ̯d/—that is, with a /d/—even though it is pronounced as [klaɪ ̯t], with a [t]. Other forms of the word have the sound [d]— the genitive form , for example: [klaɪ ̯dəs]. The pronunciation with [t] is predictable: final devoicing requires that obstruents (plosives and fricatives) be

Phonology in the Real World  225 Table 10.1 Grapheme-Phoneme correspondences for German consonants. Grapheme(s)

, (before )



, (before

and )





, (before )



Phoneme(s)

/p/ /b/ /t/ /d/ /k/ /g/ /m/ /n/ /ŋ/ /f/ /v/ /s/ /z/ /ʃ/ /ʒ/ /ç/ /ʀ/ /h/ /j/ /l/ /pf/ /ts/ /tʃ/ /dʒ/ /kv/ /ks/

Examples

,

,

, ,

, , , , ,

, , , , , ,



, ,

,

voiceless when at the end of a word or syllable. This allows us to transcribe all forms of the word phonemically with a /d/ and still arrive at the correct pronunciation of each form. The phonemes /p/, /t/, and /k/ will always be pronounced as [p], [t], and [k]. The phonemes /b/, /d/, and /g/, however, will be realized as [b], [d], and [ g] only if they occur at the beginning of a word or syllable. If they occur at the end of a word or syllable, they will be subject to final devoicing and pronounced [p], [t], and [k] (an exception is in the ending ; for more details, see the discussion below). The two /g/ phonemes in ‘against’ are pronounced [ g] because they occur at the beginning of the word and at the beginning of the second syllable: /geːgən/ [ geːgən]. The /g/ in ‘victory’, on the other hand, is pronounced [k] because it occurs at the end of the word: /ziːg/ [ziːk]. The phoneme /g/ may also be pronounced as a fricative, [ç]. If it occurs in the ending and is not followed by a vowel, it will be pronounced [ç]: ‘curly’ /lɔkɪg/ [lɔkɪç].

226  Phonology in the Real World The letters that represent the nasal phonemes in German are relatively straightforward, but you must pay attention to the position of . If it appears before a that is in the same morpheme, the combination will represent the phoneme /ŋ/: ‘rescue’ /ʀɛtʊŋ/ [ʁɛtʊŋ]. If it appears before a in the same morpheme, by itself will represent the phoneme /ŋ/: ‘tank’ /taŋk/ [taŋk]. In all other cases, will represent the phoneme /n/. In a word like ‘hand’, for example, it represents /n/: /hand/ [hant]. It also represents /n/ in word like ‘unclear’ because although appears before , it belongs to the prefix , but belongs to the root , a different morpheme: /ʊnklaːʀ/ [ʔʊnklaːɐ̯]. The letter at the beginning of a word can represent either the phoneme /f/ or the phoneme /v/. You will recall from the discussion in chapter 8 that when it appears at the beginning of a native German word like ‘father’, it is pronounced [f ]—in which case it represents the phoneme /f/: /faːtəʀ/ [faːtɐ]. When it occurs at the beginning of a word borrowed from another language, as in ‘vase’, it is pronounced [v] and represents the phoneme /v/: /vaːzə/ [vaːzə]. The letter /v/ in any other position typically represents the phoneme /v/. This is the case for in words like ‘olive’ /oːliːvə/ [ʔoliːvə] and ‘private’ /pʀiːvaːt/ [pʁivaːt]. It also holds for a that occurs at the end of nouns and adjectives. With or without an ending, the adjective ‘active’, for example, will be transcribed phonemically with /v/. When it has an ending that begins with a vowel, /v/ will be pronounced [v]: /aktiːvə/ [ʔaktiːvə]. When it does not have an ending, /v/ will occur at the end of the word and be pronounced [f ] (because of final devoicing): /aktiːv/ [ʔaktiːf ]. The letter is similar to in that it can represent more than one phoneme. A single can represent /ʃ/, /s/, or /z/. When it occurs at the beginning of a word before

or , it represents the phoneme /ʃ/: ‘mockery’ /ʃpɔt/ [ ʃpɔt]; ‘stem’ /ʃtam/ [ ʃtam]. When occurs at the beginning of a word before any consonant other than

or , you should transcribe it as the phoneme /s/: ‘slave’ /sklaːvə/ [sklaːvə]. When occurs in an ending (on a verb, noun, adjective, article), you should also transcribe it as /s/: ‘(you) come’ /kɔmst/ [kɔmst]; ‘of the dog’ /dɛs hʊndəs/ [dɛs hʊndəs]; ‘smallest’ /klaɪ ̯nst/ [klaɪ ̯nst]. In all other positions, will most often be transcribed as /z/. You should always transcribe an before a vowel at the beginning of a word as /z/: ‘son’ /zoːn/ [zoːn]. You should also transcribe a single between two vowels as /z/, as in the verb ‘to solve’ /løːzən/ [løːzən]. Notice that in all forms of the verb will be transcribed phonemically with /z/. The third person singular form, for example, ‘solves’, will be transcribed with /z/: /løːzt/ [løːst]. The /z/ in this form of the word will be pronounced [s] because of final

Phonology in the Real World  227

Transcribe a single phonemically as /ʃ/ if is at the beginning of a word before

or :   ‘spider’ /ʃpɪnə/ [ ʃpɪnə]   ‘stone’ /ʃtaɪ ̯n/ [ ʃtaɪ ̯n] /s/ if is at the beginning of a word before any other consonant or in an ending:   ‘Slav’ /slaːvə/ [slaːvə]   ‘(you) know’ /kɛnst/ [kɛnst]   ‘of the country’ /dɛs landəs/ [dɛs landəs] /z/ if is in any other position:   ‘juice’ /zaft/ [zaft]   ‘can’ /doːzə/ [doːzə]   ‘prize’ /pʀaɪ ̯z/ [pʁaɪ ̯s] Special cases: 1. Transcribe in a word stem that ends in or as /s/:   ‘service’ /diːnst/ [diːnst]   ‘kiosk’ /kiːɔsk/ [kiːɔsk] 2. Transcribe at the end of word as /s/ if it is doubled when an ending is added:   ‘secret’ /gəhaɪ ̯mnɪs/ [ gəhaɪ ̯mnɪs] ( ‘secrets’)   ‘bus’ /bʊs/ [bʊs] ( ‘busses’)

devoicing. When follows a vowel at the end of a word, it will also typically be transcribed as /z/: ‘house’ /haʊ̯z/ [haʊ̯s]. Notice that final devoicing applies to this form of the word, but not the dative form, (which occurs in the expression ‘at home’), since /z/ in the dative form occurs at the beginning of a syllable: /haʊ̯zə/ [haʊ̯zə]. Both and are used to represent the phoneme /ʒ/. This is not a common phoneme in German, so the words you encounter with and will most often represent other phonemes: typically represents /g/, as you learned above, and typically represents /j/. The list of words in which and represent /ʒ/ is not long; those that you are most likely to come across include the following: (5)

= /ʒ/ ‘passage’ (and other words ending in ), ‘jelly’, ‘genius’, ‘to be embarrassed’, ‘genre’, ‘direction’, ‘director’, ‘regime’

(6)

= /ʒ/ ‘venetian blind’, ‘jargon’, ‘jury’, ‘journal’ (and other words formed with )

228  Phonology in the Real World The combination will most often represent the phoneme /ç/. (We are of course ignoring here the use of in the combination , which represents /ʃ/.) Whether is pronounced [ç] or [x], you can typically transcribe it phonemically as /ç/. For example, you can transcribe the in ‘mug’ as /ç/: /bɛçəʀ/ [bɛçɐ]. You can also transcribe in ‘to laugh’ as /ç/: /laçən/ [laxən]. The only exception to this is the use of in the sequence , which represents /ks/ when all letters belong to a single stem: ‘salmon’ /laks/ [laks]. As you see from the words and , the phoneme /ç/ can be pronounced [ç] or [x]. It is pronounced [x] if it occurs after the phonemes /aː/, /a/, /oː/, /ɔ/, /uː/, /ʊ/, and /aʊ̯/ (see, for example, the pronunciation of /ç/ in ). In all other positions, /ç/ is pronounced [ç]. It is pronounced [ç] in , for example, because it occurs after the vowel phoneme /ɛ/. One final point that we would like to make about grapheme-phoneme correspondences involves the grapheme : You can always transcribe with the phoneme /ʀ/. It does not matter whether it has a consonantal or vocalic pronunciation. (For reasons we do not need to go into here, we use the IPA symbol for a uvular trill to represent the phoneme in German.) (7)

= /ʀ/ ‘tube’ /ʀøːʀə/ [ʁøːʁə] ‘trace’ /ʃpuːʀ/ [ ʃpuːɐ̯] ‘renter’ /miːtəʀ/ [miːtɐ]

You will need to pay attention to the position of /ʀ/ in order to determine how to pronounce it. If it occurs before a vowel in the same syllable, it will have a consonantal pronunciation, [ʁ]: ‘quick’ /ʀaʃ/ [ʁaʃ]. If it occurs after a vowel in the same syllable (any vowel other than /ə/), it will be pronounced as nonsyllablic dark schwa, [ɐ̯]: ‘me’ /miːʀ/ [miːɐ̯]. If it occurs after /ə/ in the same syllable, the combination /əʀ/ will be pronounced as dark schwa, [ɐ]: ‘liver’ /leːbəʀ/ [leːbɐ]. Although the glottal stop, [ʔ], is one of the sounds of German, there is no orthographic symbol for it and it is not considered a phoneme of German. We can predict where it will occur, so we do not need to include it in the phonemic representation of a word. We will work with this sound later in this chapter. For now, you should include it in your phonetic transcriptions in those environments in which it typically occurs: word-initially before a vowel ( ‘old’ [ʔalt]); within a compound before a word that begins with a vowel ( ‘fisheye lens’ [fɪʃʔaʊ̯gə]); and after a prefix if the following sound is a vowel ( ‘to end’ [bəʔɛndən]). The following exercises will give you practice in using German orthography to transcribe the consonant phonemes of German. You will also be asked to use the phonemic transcriptions you produce to then provide phonetic transcriptions.

Phonology in the Real World  229 Instructions for exercises 10.9 through 10.13: Supply the missing phonemes. Then transcribe each word phonetically. 10.9. /p t k/ or /b d g/? a. / ʀʊ ə/ b. / ɛʀ / c. / ɪn / d. / ɛ / e. / ə ʊl / f. / iː / g. / aʊ̯ ə/ h. / ʊʀs ɪ / i. / ɪl əʀ/ j. / aː / 10.10. /f/ or /v/? a. / aɪ ̯nd/ b. /tɔɪ ̯ əl/ c. / ɛlt/ d. /løː ə/ e. / oːgəl/ f. / œlɪg/ g. / eːnʊs/ h. / iːʀʊs/ i. /kla iːʀ/ j. / ɪktiː / k. /ɪndʊktiː ə/ 10.11. /s/ or /z/? a. /ka ə/ b. /gʀuː / c. / ɪlbə/ d. /bøː ə/ e. / yː va əʀ/ f. /gaː / g. /leː ən/ h. /laː / i. /liː t/ j. /diː ə / k. /kɛn t/ l. / oːfa /

230  Phonology in the Real World 10.12. /ʃ/, /s/, /z/, /ç/, or /k/? a. / laŋə/ b. / kɛptɪ / c. /vɛl əʀ/ d. /fʀʊ t/ e. / iː t/ f. /dɛ daɪ ̯ / g. /ɔ / h. /glaɪ ̯ / i. / aʊ̯ ən/ j. /pʀaɪ ̯ ɛŋkʊŋ/ 10.13. a. / ɔ ə / b. / aɪ ̯ ɔ c. / a ə/ d. / ɛ / e. / aʊ̯ / f. / uː ɪ / g. / ɪ ɪ ə / h. / ɛ i. / ɔ aː ə/ j. / aː ɪ k. / uː / l. / ʊ ə/ m. / ə eː / n. / ɪ ə/ o. / aɪ ̯ ə

/

ɪ / ə /

/

1 0.3.2. Vowels All the letters that represent the vowel phonemes of German are used to represent more than one phoneme. The letter , for example, can represent both /aː/ and /a/: (8)

/aː/ ‘came’ /kaːm/ [kaːm] /a/ ‘comb’ /kam/ [kam]

In chapter 8 you learned how to use information provided by consonant symbols to determine how to pronounce the vowels that precede them. You can use this same information to determine which phonemes these vowel letters represent. We can express the grapheme-phoneme correspondences for vowels with essentially the same rule of thumb that we presented in chapter 8:

Phonology in the Real World  231 (9) Rule of thumb for vowel letter and phoneme correspondences One or no consonant letter in the same morpheme indicates that the preceding vowel is long. Two or more consonant letters in the same morpheme indicate that the preceding vowel is short.

This rule of thumb accounts for the examples in (8). One consonant letter follows the in , so this vowel phoneme is long: /aː/; two consonant letters ­follow the in , so this vowel phoneme is short: /a/. The rule of thumb also accounts for the vowel grapheme-phoneme correspondences in the following examples. (10) /eː/ ‘the’ /deːn/ [deːn] /ɛ/ ‘because’ /dɛn/ [dɛn] (11)



/iː/ /ɪ/

‘style’ /ʃtiːl/ [ ʃtiːl] ‘quiet’ /ʃtɪl/ [ ʃtɪl]

Remember that endings are separate morphemes, so we do not include consonants that belong to endings when we count the consonants after the vowel in the root of a verb. The in ‘(you) live’, for example, is a long vowel phoneme because there is only one consonant letter after it in the same morpheme, . The letters belong to the verb ending, which is a different morpheme. You can take everything you learned in chapter 8 about the relationship between German orthography and the vowel sounds of the language and, with one modification (see the discussion below concerning the phonemic representation of in endings), use it to transcribe German vowels phonemically. The use of , doubled vowels, and after as additional means to indicate vowel length also applies to vowel phonemes: (12) Additional signals of length ‘bean’ /boːnə/ [boːnə] doubled vowel ‘hall’ /zaːl/ [zaːl]; ‘lake’ /zeː/ [zeː]; ‘moss’  /moːz/ [moːs] after ‘service’ /diːnst/ [diːnst]

The letter signals that a preceding vowel phoneme is long; doubled , , and represent long vowel phonemes; and immediately after signals that the is a long phoneme. The rule of thumb cannot be applied reliably to vowel phonemes in prefixes and suffixes, but you can apply what you now know about the pronunciation of the vowels in common prefixes and suffixes when you transcribe them phonemically.

232  Phonology in the Real World (13) Prefixes with short vowels before one or no consonant /a/ /ap/ [ʔap], /an/ [ʔan] /ɛ/ /ɛʀ/ [ʔɛɐ̯], /fɛʀ/ [fɛɐ̯], /tsɛʀ/ [tsɛɐ̯] /ə/ /bə/ [bə], /gə/ [ gə] /ʊ/ /ʊm/ [ʔʊm], /ʊn/ [ʔʊn] (14) Suffixes with short vowels before single consonants /ɪg/ [ɪç], /ɪn/ [ɪn], /nɪs/ [nɪs] /ɪ/ /ə/ /çən/ [çən]

As these examples demonstrate, you can transcribe the vowel phonemes in these affixes using the same symbols that indicate how they are pronounced. You will be happy to learn that the phonemic representation of at the end of a word or in an ending is completely straightforward: at the end of a word or in an ending represents the phoneme /ə/. (15) /ə/

‘wool’ /vɔlə/ [vɔlə], ‘(I) buy’ /kaʊ̯fə/ [kaʊ̯fə],   ‘works’ /aʀbaɪ ̯tət/ [ʔaɐ̯baɪ ̯tət], ‘to threaten’ /dʀoːən/ [dʁoːən], ‘more beautiful’ /ʃøːnəʀ/ [ ʃøːnɐ], ‘more beautiful’ /ʃøːnəʀə/ [ ʃøːnəʁə]

Notice that although we pronounce at the end of a word as dark schwa, [ɐ], we do not treat dark schwa as a phoneme of German. Instead, we represent it phonemically as a sequence of phonemes: /əʀ/. The word ‘more beautiful’, for example, is transcribed phonemically as /ʃøːnəʀ/, with transcribed as /əʀ/, but it is pronounced [ ʃøːnɐ], with [ɐ] as the phonetic realization of /əʀ/. When this word has an ending like , we still transcribe the sequence as /əʀ/, but we pronounce it [əʁ]: ‘more beautiful’ /ʃøːnəʀə/ [ ʃøːnəʁə]. Because we can predict when the sequence /əʀ/ will be pronounced [ɐ] and when it will be pronounced [əʁ], we can represent both pronunciations with the sequence /əʀ/. (You will recall from the discussion of /ʀ/ above that /əʀ/ is pronounced [ɐ] when /ʀ/ follows /ə/ in the same syllable; when /ʀ/ precedes a vowel in its syllable, as it does in the word /ʃøːnəʀə/, it is pronounced [ʁ].) Table 10.2 lists the grapheme-phoneme correspondences for the vowel phonemes of German and includes diphthongs as well as monophthongs. If you use this table together with the information provided above—the rule of thumb in (9); additional signals of length; vowel phonemes in prefixes and suffixes—you will be able to use German spelling to determine the phonemic representation of the words you commonly encounter in German. The pronunciation of the vowel phonemes in German is straightforward. You can pronounce all the short vowels and diphthongs just as you might expect. For example, /ɪ/ is pronounced [ɪ], /ɛ/ is pronounced [ɛ], /aɪ ̯/ is pronounced [aɪ ̯], and so on.

Phonology in the Real World  233 Table 10.2 Grapheme-Phoneme correspondences for German vowels. Grapheme(s)









Phoneme

/aː/ /a/ /eː/ /ɛ/ /ə/ /iː/ /ɪ/ /oː/ /ɔ/ /uː/ /ʊ/ /eː/ /ɛ/ /øː/ /œ/ /yː/ /ʏ/ /yː/ /ʏ/ /aɪ ̯/ /aʊ̯/ /ɔɪ ̯/

Examples













, , ,

,

The long vowel phonemes are a bit different, because there are two ways you can pronounce each of them. When long vowel phonemes occur in stressed syllables (syllables that are louder and last longer than other syllables), they are pronounced as long vowels. If they do not occur in a stressed syllable, they are short. They do not change in quality—only in quantity (duration). All the vowels in the word ‘biology’, for example, are long vowel phonemes, /biːoːloːˈgiː/, but only the last vowel is pronounced as a long vowel because it is the only one that occurs in a stressed syllable (marked with the diacritic [ˈ]): [bioloˈgiː]. The other vowels are still tense, but short. The following exercises will give you practice in using German orthography to transcribe the vowel phonemes of German. You will also be asked to use the phonemic transcriptions you produce to then provide phonetic transcriptions. In words with more than one syllable, we will use the diacritic [ˈ] to indicate which syllable receives primary stress. You will need to use this information to determine the pronunciation of long vowel phonemes. If a long vowel phoneme is not in a syllable with primary stress, you will need to transcribe it phonetically as a short vowel.

234  Phonology in the Real World Instructions for exercises 10.14 through 10.16: Supply the missing phonemes. Then transcribe each word phonetically. 10.14. a. /ˈd ŋk / b. /ˈj ʀg ŋ/ c. /ˈgʀ s n/ d. /ˈv nʃ n/ e. /ˈl n / f. /ˈts l / g. /v bt/ h. /g lb/ i. /ˈʀ k st/ j. /ˈv ʀtʃv l/ 10.15. Word with prefixes and suffixes. a. /ˈb s g/ b. /ˈp ʀç n/ c. /ˈd m ʀ ŋ/ d. /f ʀˈm g n/ e. /g ˈz ndh t/ f. /ˈk nd ʃ/ g. /ˈp ŋktl çk t/ h. /g ˈf ŋn s/ i. /ts ʀˈʀ s n/ j. /ˈ nʃ ld/ 10.16. Shortening of long vowel phonemes. Keep in mind when you transcribe these words phonetically that long vowel phonemes are shortened if they occur in an unstressed syllable. a. /m l ˈd / b. /ˈf t / c. /ʀ ˈn / d. /p ˈʀ t/ e. /f t ˈgʀ f/ f. /ʀ ˈ l/ g. /ʀ l ˈt t/ h. /p ˈl ʀt/ i. /p ˈp l / j. /b l ˈg / The following exercises provide you with the opportunity to transcribe entire words phonemically and phonetically.

Phonology in the Real World  235 10.17. Transcribe the following words phonemically. Then transcribe each word phonetically. a. b. c. d. e. f. g. h. i. j. k. l. m. n. o. p. q. r. s. t. 10.18. Transcribe the words in table 10.1 phonemically and phonetically. 10.19. Transcribe the words in table 10.2 phonemically and phonetically. 1 0 .4. Pronunciation Practice II In this chapter we focus on those sounds in German that are similar in some way to sounds in English. They may share some phonetic features with sounds in English; they may be represented with the same letter. Sounds that are similar in two languages typically pose the greatest difficulty for learners. When two sounds are similar, learners may not perceive the differences between them and simply substitute the sound from their native language for the sound in the target language—the language they are attempting to learn. Even though this may not cause miscommunication, the resulting pronunciation will not be native-like. Our goals here are to make you aware of the differences between pairs of similar sounds in German and English and to help you fine-tune your pronunciation. Our discussion of consonants will focus on the pronunciation of and . We will also work with the glottal stop, [ʔ]. Our discussion of vowels will include the back vowels, [uː ʊ oː ɔ], and [iː] and [eː]; we will also discuss the pronunciation of vowels in unstressed syllables.

236  Phonology in the Real World 1 0.4.1. Similar German Consonants In chapter 8 we worked with the vocalic pronunciation of , namely dark schwa. You learned to use the syllabic pronunciation of dark schwa for the ending in a word like ‘littler’ [klaɪ ̯nɐ] and the nonsyllabic pronunciation for at the end of a syllable in a word like ‘me’ [miːɐ̯]. In this chapter we will focus on consonantal , the pronunciation of at the beginning of a syllable. There are essentially three ways to pronounce consonantal , all of which are acceptable: [r], [ʀ], and [ʁ]. We will describe how each of these sounds is produced and provide some tips on learning how to articulate them. You should choose one of these sounds (the one you find easiest to approximate) and aim to use it consistently for consonantal . The alveolar trill, [r], also called an apical trill, is used primarily by speakers in Bavaria and Austria. It is produced by causing the tip of the tongue (the apex— hence the name “apical”) to vibrate against the alveolar ridge. This is a voiced sound, so the vocal cords are also vibrating while it is articulated. Some speakers have no difficulty producing this sound. If you have had some exposure to Spanish, it may help you to know that it is the r-sound in the Spanish word for ‘dog’, . To learn to make this sound, place the tip of your tongue, very loosely, against the alveolar ridge, just behind the upper front teeth. Take a deep breath and then breathe out with a good deal of force, keeping the tip of your tongue relaxed so that it flaps against the alveolar ridge. You will probably need to adjust the tension of your tongue in order to set it in vibration with the current of air you are producing. This may take some work. Once you are able to produce the vibration necessary to say this sound, you will need to make sure that it is voiced. If you hold several fingers on your throat while you are saying it and do not feel any vibration, the sound you are producing is a voiceless apical trill, [r̥] (the diacritic [˳] indicates that a sound is voiceless). You can add voicing by doing exactly what you do when you say an [s] and then switch to a [z]. Experience what it is like to say [ssszzz] and then work on doing this with the apical trill: [r̥r̥r̥rrr]. The final step is to work on saying [r] in words in various environments: at the beginning of a word before a vowel ( ‘speech’), after a word-initial consonant ( ‘to print’), word-internally between two vowels ( ‘berry’), and so on. The uvular trill, [ʀ], is produced when the uvula vibrates against the back of the tongue. You may be able to cause your uvula to vibrate and produce a uvular trill by first gargling without any liquid. Some students report success when they attempt this first thing in the morning. This is not an easy sound to produce (very few languages use a uvular trill; see Ladefoged 2005:166), so you should not be discouraged if you cannot set your uvula in motion and produce the right kind of

Phonology in the Real World  237 vibrations for a trill. Chances are very good that you are producing a voiced uvular fricative, [ʁ], the more common pronunciation of consonantal . The voiced uvular fricative, [ʁ], can be viewed as the voiced counterpart to an [x] that is produced further back in the mouth. The symbol for this sound, a voiceless uvular fricative, is [χ]. The [ʁ] sound is produced by raising the back of the tongue toward the uvula or the back of the velum and forming a narrow passage so that turbulence is created when an airstream is forced through. This sound is voiced, so the vocal cords are vibrating while it is being articulated. If you can produce [x], you can use this as a starting point for learning to pronounce [ʁ]. First say the nonsense word [ʔaːxaː]. Then move the back of your tongue further back in your mouth when you pronounce the , and you will be saying [ʔaːχaː]. The final step is to voice the [χ], and you will then be saying [ʔaːʁaː]. After you can comfortably produce [ʁ] in this context, practice it without a preceding vowel by saying [ʁaː]. If you can then produce it before all the vowels in German ([ʁeː], [ʁiː], [ʁoː], and so on), you will be ready to use it for all instances of consonantal . Because there are three acceptable ways to pronounce consonantal in German, you should be able to find one that you can produce and thus avoid using a North American , which is an alveolar approximant, [ɹ], a sound that is made with no contact between the articulators. If you attempt to make an alveolar trill, [r], and find yourself saying a North American , you should focus on making contact with the tip of your tongue and the alveolar ridge. Even if you can only manage one strike with your tongue against the alveolar ridge, this is an acceptable pronunciation of consonantal . It is a sound called a flap, [ɾ], and German speakers who use an alveolar trill will use an alveolar flap as a reduced (less strongly articulated) form of consonantal . Speakers of North American En­ glish use this sound for in a word like , so another strategy for avoiding a North American would be to aim for this sound, which you should have no trouble making. German is the lateral approximant, [l], which is produced by placing the tip of the tongue against the alveolar ridge and allowing air to escape on both sides of the tongue (while the vocal cords vibrate). In English, most speakers have two kinds of sounds. One is like German , which is also called “clear .” In English, [l] is typically the pronunciation for at the beginning of a syllable before a vowel and before the sound [ j], as in words like , , , and . The other sound is a velarized , [ɫ], which is also called “dark .” This sound is produced essentially like [l], but the back of the tongue is also raised toward the velum. This is the sound that North Americans typically use at the end of syllables after vowels, as in , , , and . The Standard German is always a clear , so it is important for you to concentrate on producing a clear in all contexts.

238  Phonology in the Real World It will be helpful to learn to hear the difference between clear and dark before you practice making clear in environments that are new for you. Exercise 10.26 below should help you hear the difference. To learn to use a clear at the end of a syllable after a vowel, start by saying words like the following, which have an at the beginning of a syllable before a vowel (we use the symbol [.] to show where the syllable boundaries in these words are located). (16)

‘villa’ [vɪ.la] ‘pullover’ [pʊ.li] ‘polo’ [poː.lo] ‘aluminum’ [ʔaː.lu]

You should be able to pronounce the in these words with a clear because they are in a position where speakers of English typically use a clear . Now say these words again, but hold the [l] so that you can feel the position of your tongue while you are producing a clear : , , , . The next step is to leave off the final vowel, making sure that you keep your tongue in position for the clear . You will end up saying the following words, and you should be pronouncing them with a clear : ‘wants’ [vɪl], ‘row’ [pʊl], ‘pole’ [poːl], ‘eel’ [ʔaːl]. You can then add some complexity and a bit more difficulty by adding a consonant after an in this position. Practice saying words like ‘wild’, ‘desk’, ‘polarizes’, and ‘paints’. The final similar consonant we will work with is the glottal stop, [ʔ]. This is not a phoneme of German because it is not capable of causing a difference in meaning. If you were to pronounce ‘office’ as [amt] rather than [ʔamt], for example, this would not result in a different word or bring about any confusion. However, if you do not use glottal stops as they are typically used by native speakers of German, your speech will lack a characteristic feature of the language. Glottal stops are found word-initially before a vowel; within a compound before a word that begins with a vowel; and after a prefix if the following sound is a vowel. (17) ‘corner’ [ʔɛkə] ‘street corner’ [ ʃtʁaːsənʔɛkə] ‘opened’ [ gəʔœfnət]

Although speakers will omit the glottal stop in quick, relaxed speech, it is a feature of German that you should learn because it is so characteristic of German pronunciation. 1 0.4.1.1. Perceptual Practice 1 The perceptual exercises here and the production exercises below focus on [ʁ] as the consonantal pronunciation of . This is the most common pronunciation,

Phonology in the Real World  239 and it is also probably the easiest to learn for native speakers of North American English. 10.20. North American [ɹ] versus German [ʁ]. Listen to the following pairs of words. The first word in each pair is an English word with [ɹ]; the second is a similar German word, with [ʁ]. a.

b.

c. d.

e. f. g. h. 10.21. North American [ɹ] versus German [ɐ̯]. Listen to the following pairs of words. The first word in each pair is an English word with [ɹ]; the second is a similar German word, with [ɐ̯]. a.

b.

c.

d.

e.

10.22. [x] or [ʁ]? Consider each pair of words and circle the word you hear. Some of these are real words. Others are “nonsense” words; they are possible words of German, but do not actually exist. a. [ˈmaʁə] [ˈmaxə] b. [ˈdɔʁo] [ˈdɔxo] c. [ˈfluːʁən] [ˈfluːxən] d. [ˈʃaʁəl] [ˈʃaxəl] e. [ˈvɔʁə] [ˈvɔxə] f. [ˈnaʁə] [ˈnaxə] g. [ˈkuːʁən] [ˈkuːxən] h. [ˈmaʁi] [ˈmaxi] 10.23. Listen to the following words and write which phone you hear for each /ʀ/, [ʁ] or [ɐ̯]. Remember that /ʀ/ before a vowel in the same syllable is pronounced [ʁ]; /ʀ/ after a vowel in the same syllable (any vowel other than /ə/) is pronounced [ɐ̯]. a. /ʀeːdə/ b. /uːʀtaɪ ̯l/ c. /naːʀʊŋ/

240  Phonology in the Real World d. /dʀʊk/ e. /ʃpatsiːʀən/ f. /ʀaʊ̯btiːʀ/ g. /knaʀən/ h. /leːʀlaʊ̯f/ 10.24. Listen to the following idiomatic expressions and write the phonetic symbol(s) for the sound(s) you hear for the letters that are underlined. a. wie Kraut und Rüben b. Ross und Reiter nennen c. einen über den Durst trinken d. etwas durch eine gefärbte Brille betrachten e. vom Regen in die Traufe geraten f. Probieren geht über Studieren. g. die Werbetrommel für etwas rühren h. einen Verbrecher den Händen der Gerechtigkeit übergeben 10.25. Listen to the following text and fill in the blanks with [ʁ], [ɐ̯], or [ɐ]. Vo[ ] dem Gesetz Vo[ ] dem Gesetz steht ein Tü[ ]hüt[ ]. Zu diesem Tü[ ]hüt[ ] kommt ein Mann vom Lande und bittet um Eint[ ]itt in das Gesetz. Ab[ ] de[ ] Tü[ ]hüt[ ] sagt, daß e[ ] ihm jetzt den Eint[ ]itt nicht gewäh[ ]en könne. De[ ] Mann üb[ ]legt und f[ ]agt dann, ob e[ ] also spät[ ] we[ ]de eint[ ]eten dü[ ]fen. »Es ist möglich,« sagt de[ ] Tü[ ]hüt[ ], »jetzt ab[ ] nicht.« Da das To[ ] zum Gesetz offensteht wie imm[ ] und de[ ] Tü[ ]hüt[ ] beiseite t[ ]itt, bückt sich de[ ] Mann, um du[ ]ch das To[ ] in das Inne[ ]e zu sehn. Als de[ ] Tü[ ]hüt[ ] das me[ ]kt, lacht e[ ] und sagt: »Wenn es dich so lockt, ve[ ]suche es doch, t[ ]otz meines Ve[ ]botes hineinzugehn. Me[ ]ke ab[ ]: Ich bin mächtig. Und ich bin nu[ ] de[ ] unt[ ]ste Tü[ ]hüt[ ]. Von Saal zu Saal stehn ab[ ] Tü[ ]hüt[ ], ein[ ] mächtig[ ] als de[ ] ande[ ]e. Schon den Anblick des d[ ]itten kann nicht einmal ich meh[ ] e[ ]t[ ]agen.« Franz Kafka (1919)

10.26. Dark versus clear . Listen to the following pairs of words. The first word in each pair is an English word with [ɫ]; the second is a similar German word, with [l]. a.

b.

c.

d.

e.

f.

g.

Phonology in the Real World  241 h. i. j. k.



10.27. [ʔ] or [h]? Listen to the following words and circle the one you hear. Remember that a glottal stop is pronounced before a vowel at the beginning of a word; within a compound before a word that begins with a vowel; and after a prefix if the following sound is a vowel. a.

b.

c.

d.

e.

f.

g.

h.

10.28. With [ʔ] or without? Listen to the following phrases and underline the one you hear. We use the symbol “#” for internal word boundaries so that you can more easily visualize the differences between some of the pairs below. a.

b.

c.

d.

e.

f.

g.

h.

1 0 .4.1.2. Production Practice 1 The following exercises are designed to help you improve and fine-tune your pronunciation of /ʀ/ and /l/ and to more automatically produce the glottal stop, [ʔ], in those environments in German in which it is typically used. 10.29. Practice pronouncing the words in the right-hand column in exercise 10.20. Remember to use a consonantal pronunciation for in these words—either [r], [ʀ], or [ʁ]. 10.30. Practice pronouncing the words in the right-hand column in exercise 10.21. Remember to pronounce in these words as [ɐ̯]. 10.31. Listen to and repeat the words in exercise 10.23, focusing on your pronunciation of consonantal and vocalic /ʀ/.

242  Phonology in the Real World 10.32. Listen again to the expressions in exercise 10.24. Then practice saying them, concentrating on your pronunciation of the segments that are underlined. 10.33. Listen again to the text in exercise 10.25. Then practice reading it aloud, paying close attention to your pronunciation of all the sounds. 10.34. Listen to the following pairs of words and repeat them, making sure to keep the back of your tongue low when you pronounce . a.

b.

c. d.

e.

f.

g.

h.

10.35. Practice saying the words in the right-hand column in exercise 10.26, making sure to produce a clear by keeping the back of your tongue low. 10.36. Listen to the following expressions, paying attention to the clear quality of each [l]. Then say each expression yourself, focusing on your pronunciation of [l]. a. Morgenstund hat Gold im Mund. b. Auf gar keinen Fall! c. Wer die Wahl hat, hat die Qual. d. Das ist halb so wild! e. Er nimmt kein Blatt vor den Mund. f. Sobald die Maus satt ist, schmeckt das Mehl bitter. g. Das ist viel Holz! h. Ich fühle mich wie im falschen Film. 10.37. Practice saying the words and phrases in the left-hand column in exercise 10.27. Remember to use a glottal stop, [ʔ], before a vowel at the beginning of a word. 10.38. Practice saying the words and phrases in the left-hand column in exercise 10.28. Remember to use a glottal stop, [ʔ], before a vowel at the beginning of a word; within a compound before a word that begins with a vowel; and after a prefix if the following sound is a vowel. 10.39. Say each infinitive and then form and pronounce each past participle. Example: a. b.

Phonology in the Real World  243 c. d. e. f. g. h. i. j. 10.40. Pronounce the following phrases and sayings, taking care to use the glottal stop in those environments in which it is typically used. Pay attention to your pronunciation of and as well. a. Aller Anfang ist schwer. b. Ehre, wem Ehre gebührt! c. Aus den Augen, aus dem Sinn. d. einen Eid ablegen e. wie auf Eiern gehen f. unter allen Umständen g. In Ulm, um Ulm und um Ulm herum. h. Ende gut, alles gut. 10.41. Here are the words to a well-known folksong in German. Before you practice saying it, underline all the vowels that you should pronounce with a preceding glottal stop. Heut kommt der Hans zu mir Heut kommt der Hans zu mir freut sich die Lies. Ob er aber über Oberammergau oder aber über Unterammergau oder aber überhaupt nicht kommt ist nicht gewiss.

1 0 .4.2. Similar German Vowels Four of the vowels in German that are similar to vowels in English are long and tense: [iː], [uː], [eː], and [oː]. In the perception exercises below, you will have a chance to compare each of these sounds to their English counterparts. The points of difference you will want to become aware of are the following: The German vowels are long, tense monophthongs, whereas their English counterparts tend to be diphthongized, especially when they occur at the end of a word. Compare, for example, German [ geː] and English [ geɪ ̯]. As Moulton (1962:97) notes, the German sounds seem to be exaggerated in comparison to their English

244  Phonology in the Real World counterparts. German [iː] is articulated with the tongue very high and very far forward in the mouth and with the lips very spread; [uː] is produced with the tongue very high and very far back in the mouth, with the lips very rounded. Both [iː] and [uː] are very tense and monophthongal; the tongue does not move while they are being articulated. These same characteristics apply to the mid vowels [eː] and [oː]: they are somewhat higher than their English counterparts; [eː] is further forward and [oː] is further back; and the lips are very spread for [eː] and very rounded in the case of [oː]. When you practice these vowels, be careful not to move your tongue when you are articulating them; you want to prevent them from becoming diphthongs, as in English. Exaggerate your lip position: spread your lips tightly for [iː] and [eː] and protrude your lips clearly when you round them for [uː] and [oː]. Make sure that your tongue height is a bit higher for each sound than its counterpart in English. Your tongue should also be further forward for [iː] and [eː] than for their English counterparts and further back for [uː] and [oː]. The two remaining back vowels in German, [ʊ] and [ɔ], are short and lax and therefore not diphthongized by native speakers of North American English. German [ʊ] is articulated with more rounding than its English counterpart (the vowel in words like and ), so you should make sure you have enough lip rounding when you fine-tune your pronunciation of this sound. Because often represents [ʌ] in English (a vowel similar to dark schwa, [ɐ]), you may also be tempted to pronounce German [ʊ] as [ʌ]. Compare English [bʌt] and German ‘flounder’ [bʊt]. If you find yourself doing this when you read German, just remember that can only represent [ʊ] (or [uː]), never a vowel like English [ʌ]. The sound [ɔ] is a bit more complicated because of the regional variation in North American English. In the perception exercises below we will contrast the [ɑː] in English (as in ) and the [ɔː] in English with the [ɔ] in German ‘buck’. Some speakers of English do not make a distinction between the vowels in and ; these speakers use [ɑː] for both. If you are one of these speakers, or if you catch yourself saying [ɑː] instead of German [ɔ], you will need to close your mouth a bit (which raises your tongue), round your lips, and make sure that the vowel you produce is short: [ɑː] is a low back long unrounded vowel, whereas [ɔ] is a mid back short rounded vowel. If you are a speaker who uses [ɔː] for , your English [ɔː] will be closer to German [ɔ], but you will need to fine-tune it a bit. German [ɔ] is a bit higher (you will need to close your mouth a bit) and shorter. The contrastive perception exercises below should help you hear the differences between English and German and help you better approximate German [ɔ]. When the long vowels of German occur in unstressed syllables, they are shortened. In a noun like ‘topic’, the is a long tense vowel, [ˈteːma], but in

Phonology in the Real World  245 the related adjective ‘thematic’, it is short because it is unstressed— but it is still a tense vowel: [teˈmaːtɪʃ]. In English, there are also pairs of related words like this that have these same kinds of stress shifts: and , and , and , and so on. English differs from German in that when a vowel in these pairs shifts from stressed to unstressed, it is typically pronounced as [ə] or [ɪ]. The [æ] in becomes [ə] in ; the [iː] in becomes [ɪ] in . You may have a tendency to do something similar when you speak German. That is, you may have a tendency to pronounce unstressed vowels as schwa. For example, instead of saying [teˈmaːtɪʃ], you may say *[təˈmaːtɪʃ] (the asterisk indicates that this is not a correct pronunciation). You may also pronounce a word like ‘sofa’ as *[ˈzoːfə] instead of [ˈzoːfa]. The perceptual exercises below will help you hear the difference between the unstressed vowel in a word like North American English and its counterpart in the very similar German . The production exercises will give you the opportunity to focus on using full vowels in unstressed syllables instead of pronouncing them as schwa. 1 0 .4.2.1. Perceptual Practice 2 The first step in acquiring a more native-like pronunciation of German vowels is to be able to hear the difference between these vowels and those in English that are most like them. If you can hear the difference between the vowels in English and German ‘cattle’, for example, you will be on your way to producing a more native-like [iː]. The following exercises will give you the opportunity to listen to the differences that have been identified in this chapter. 10.42. Listen to the following pairs of words. The first word in each pair is an En­ glish word with a vowel that is diphthongal; the second is a similar German word, with a long tense monophthong. a.

b.

c.

d.

e.

f. g.

h.

i. j.

k.

l.

246  Phonology in the Real World 10.43. Listen to the preface from “Max und Moritz,” a tale by Wilhelm Busch (1970) told in rhymed couplets. Pay attention to the pronunciation of the underlined segments, which are the long tense monophthongs [iː uː eː oː]. Max und Moritz: Eine Bubengeschichte in Sieben Streichen von Wilhelm Busch Vorwort Ach, was muß man oft von bösen Kindern hören oder lesen! Wie zum Beispiel hier von diesen, Welche Max und Moritz hießen; Die, anstatt durch weise Lehren Sich zum Guten zu bekehren, Oftmals noch darüber lachten Und sich heimlich lustig machten. Ja, zur Übeltätigkeit, Ja, dazu ist man bereit! Menschen necken, Tiere quälen! Äpfel, Birnen, Zwetschgen stehlen Das ist freilich angenehmer Und dazu auch viel bequemer, Als in Kirche oder Schule Festzusitzen auf dem Stuhle. Aber wehe, wehe, wehe! Wenn ich auf das Ende sehe!! Ach, das war ein schlimmes Ding, Wie es Max und Moritz ging! Drum ist hier, was sie getrieben, Abgemalt und aufgeschrieben.

10.44. English [ʌ] versus German [ʊ]. The following pairs of English and German words contain the symbol . This symbol represents [ʌ] in the English words and [ʊ] in those in German. Listen to the contrast in these two vowels. a.

b.

c.

d.

e. f.

g. h.

Phonology in the Real World  247 10.45. English [ɑː] versus German [ɔ]. The English words in each pair contain the vowel [ɑː]; those in German contain [ɔ]. Listen to the contrast. a.

b.

c.

d.

e. f. g.

h. 10.46. English [ɔː] versus German [ɔ]. German [ɔ] is a bit higher and shorter than English [ɔː]. Listen to the difference. a.

b.

c.

d.

e. f. g.

h.

10.47. Unstressed vowels. The underlined vowels in the English words below are unstressed and pronounced [ə]. Their counterparts in the corresponding words in German are also unstressed, but they are “full vowels” (vowels other than [ə]). Listen to the difference. a. [ˈzoːfa] b. [ˈkɔma] c. [ˈmiːnimʊm] d. [ʔakaˈdeːmɪʃ] e. [filoˈzoːfɪʃ] f. [ˈbɪʃɔf ] g. [dʁaˈmaːtɪʃ] h. [meˈtoːdɪʃ] 1 0 .4.2.2. Production Practice 2 10.48. Practice saying the words in the right-hand column in exercise 10.42. Remember not to move your tongue while you are pronouncing the vowels in these words, [iː uː eː oː]. 10.49. Listen again to the preface to “Max und Moritz” in exercise 10.43, then practice saying the text aloud, remembering to pronounce the vowels that are underlined as long, tense monophthongs.

248  Phonology in the Real World 10.50. Practice saying the words in the right-hand column in exercise 10.44. 10.51. Pronounce the following phrases and sayings, paying special attention to your pronunciation of [ʊ]. a. etwas Punkt für Punkt besprechen b. keinen Schuss Pulver wert sein c. ein Sprung ins Ungewisse d. bekannt sein wie ein bunter Hund e. etwas auf die leichte Schulter nehmen f. Es ist alles in Butter! g. Mit Geduld und Spucke fängt man eine Mucke! h. kurz angebunden sein 10.52. Practice pronouncing the words in the right-hand column in exercises 10.45 and 10.46. 10.53. Pronounce the following phrases and sayings, paying special attention to your pronunciation of [ɔ]. a. jemanden den Kopf kosten b. einen Korb bekommen c. von seinem hohen Ross herunterkommen d. jemandem ein Dorn im Auge sein e. Die Sonne bringt es an den Tag. f. Wie gewonnen, so zerronnen. g. Was du heute kannst besorgen, das verschiebe nicht auf morgen. h. Da ist Hopfen und Malz verloren. 10.54. Practice saying the following lines from poems by Johann Wolfgang von Goethe. Pay special attention to your pronunciation of [ɔ]. a. So rennet nun alles in vollem Galopp Und kürt sich im Saale sein Plätzchen; Zum Drehen und Walzen und lustigen Hopp Erkieset sich jeder ein Schätzchen. (“Hochzeitslied,” Goethe 1978) b. Das Wasser rauscht’, das Wasser schwoll, Ein Fischer saß daran, Sah nach dem Angel ruhevoll, Kühl bis ans Herz hinan. (“Der Fischer,” Goethe 1978) c.

Ich wollt’, ich wäre Gold, Dir immer im Sold; Und tätst du was kaufen, Käm’ ich wieder gelaufen. Ich wollt’, ich wäre Gold, Dir immer im Sold. (“Liebhaber in allen Gestalten,” Goethe 1949)

Phonology in the Real World  249 10.55. Practice pronouncing the words in the right-hand column in exercise 10.47. Remember that none of these words contains [ə]. 10.56. Pronounce the following phrases, taking care to use “full vowels” (not [ə]) for the vowels that are underlined. a. Giraffen und Elefanten b. Weltproduktion von Bananen c. atmosphärische Grenzschicht d. philosophische Anthropologie e. Akademisches Gymnasium Innsbruck f. Kolumbus entdeckte Amerika g. automatische Parallelen ziehen h. Makkaroni mit Käse i. alkoholische Getränke j. das garantierte Minimum 1 0 .5. Acoustic Analysis: Similarities between German and English The exercises in this chapter have provided you with an opportunity to hone your perception and production of consonants and vowels that are similar in German and North American English. By carrying out the following acoustic analyses in Praat, you will be able to observe some of the most important acoustic correlates of the differences we have described. In the exercises in this section, you will have the opportunity to compare German and North American English , consonantal , glottal stops, and the similar vowels , , , and . 1 0 .5.1. Analyzing Similar Consonants As we noted above, the main difference between German and North American English is that North American English speakers typically produce two positionally determined variants: “clear ,” [l], and “dark ,” [ɫ]. German speakers only produce clear , regardless of where it occurs. Since word-final position is one of the positions in which North American English speakers produce the dark , it is this position that we will focus on in the analyses. Compare the productions in German ‘bright’ and English in figure 10.1. The first and second formants have been marked so that you can compare the productions in terms of one important difference. You will notice that F1 and F2 are further apart in the production of the clear German [l] than they are in the production of the dark North American English [ɫ]. Both sounds have low F1 values because both are pronounced by raising the tip of the tongue and touching the alveolar ridge. You may remember from chapter 9 that a high tongue position results in lower F1 values than a low tongue position. The F2 values for the German clear [l] are higher than for the dark [ɫ], however, because [ɫ] is produced by also raising the back of the tongue toward the velum, which lowers the F2 value. Keep in mind

250  Phonology in the Real World

Figure 10.1 [l] in German ‘bright’ and [ɫ] in American English with first and second formants marked.

that it is not the absolute formant values, but rather the wider distribution of the formants, that matters in the production of clear German [l] as compared to North American English [ɫ]. 10.57. Record the German words and , paying special attention to your production of in both words. Now record the English words and . Listen to all of the words, and determine whether you have produced a clear or dark in each instance. Open the spectrograms of the German words and as well as the English words and . Now examine your productions. In the table below, enter the mean F1 and F2 values for every instance of in the words. To what extent do the formant values correspond to or differ from your expectations? To what extent do the values differ depending on the position of the in the word (syllable-initial or syllable-final)? Why might this be? What can you do to ensure that you produce a clear German in all instances? You may wish to re-record yourself to determine whether this strategy works. Remember that your goal is not to produce a spectrogram for German [l] that looks exactly like the one in figure 10.1. You should be aiming to produce German [l] with F1 and F2 values that are further apart than those of North American English [ɫ].

F1

F2

German