Word order choice, expression of 
information status and global intonation 

patterns in heritage Russian 
 

Von der Philosophisch-Historischen Fakultät der Universität 

Stuttgart zur Erlangung der Würde einer Doktorin der 

Philosophie (Dr. phil.) genehmigte Abhandlung 
 

Vorgelegt von 

Yulia Zuban 
aus Kemerowo 

 

 

   

 Hauptberichterin:  Prof. Dr. Sabine Zerbian 

 Mitberichter:  Prof. Dr. Luka Szucsich 

   

   

 Tag der mündlichen Prüfung:  17.05.2024 

 

 

 

 

 

 

Institut für Linguistik / Anglistik 

der Universität Stuttgart 

2024  



ii 

Erklärung über die Eigenständigkeit der Dissertation 

Ich versichere, dass ich die vorliegende Arbeit mit dem Titel “Wortstellung, Informationsstatus 

und Intonation im Russischen als heritage Sprache” selbständig verfasst und keine anderen als die 

angegebenen Quellen und Hilfsmittel benutzt habe; aus fremden Quellen entnommene Passagen 

und Gedanken sind als solche kenntlich gemacht. 

 

Declaration of Authorship 

I hereby certify that the dissertation entitled “Word order choice, expression of information 

status and global intonation patterns in heritage Russian” is entirely my own work except where 

otherwise indicated. Passages and ideas from other sources have been clearly indicated. 

 

Name/Name:  Yulia Zuban 

Unterschrift/Signed:  

Datum/Date:   



iii 

Acknowledgments 

This thesis could not have been written without the support of many people, all of whom I will 

probably not be able to mention here. 

Most of all, I would like to thank my “Doktormutter” Sabine Zerbian and my “Doktorfater” 

Luka Szucsich for their support with this thesis. Sabine, thank you for your inspiration, 

constructive feedback, and long conversations about linguistic topics related to my thesis and 

beyond. Luka, thank you for transmitting your deep knowledge about topics related to syntax, not 

only in Russian but in many other languages. 

My warmest thanks and gratitude go to all the members of the RUEG research unit ("Emerging 

Grammars in Language Contact Situations: A Comparative Approach") who contributed 

enormously to the development of the ideas of this thesis and who have become my linguistic 

family for the last six years. Dear PIs of RUEG, Natalia Gagarina, Shanley E. M. Allen, Christoph 

Schroeder, Heike Wiese, Artemis Alexiadou, Rosemarie Tracy, Anke Lüdeling, Oliver Bunk, 

Mareike Keller, Anna Shadrova and Judith Purkarthofer, thank you all for your support at different 

levels of research, conceptual and practical.  

I would also like to thank the Mercator fellows of RUEG, Maria Polinsky, Shana Poplack, 

Cristina Flores, and Jeanine Treffers-Daller, for their inspiration and their ability to see the bigger 

picture of heritage languages and their speakers. My special thank you goes to Tamara Rathcke 

for her expertise on Russian prosody. Thank you, Esther Jahns, Pia Linscheid, and İrem Duman 

Çakır for the coordination of the RUEG project. 

Dear PhDs of RUEG, Wintai Tsehaye, Vicky Rizou, Kateryna Iefremenko, Martin Klotz, 

Annika Labrenz, Kalliopi Katsika, Onur Özsoy, Nadine Zürn, Johanna Tausch, Sofia Grigoriadou 

and Kristina Barabashova, thank you for your expertise on different languages and phenomena 

and your life expertise on how to survive writing a PhD thesis. 

My special and warm спасибо goes to Tatiana Pashkova who helped me with data collection 

in the US and who discussed big questions of life and research with me. Another special спасибо 

is dedicated to Maria Martynova who was my frequent partner in writing articles and analyzing 

data. 

I cannot mention all of the research assistants who worked in RUEG at different times, but I 

would like to thank them all for their great help with data transcription and annotation. My special 

большое спасибо goes to Yuliia Ivaschyk, Alina Schoepf and Mariia Perfilova, за многочасовые 

аннотации данных. Yuliia Ivaschyk, thank you very much for the data collection in the US 



Acknowledgements 

iv 

despite all the obstacles on our way and weather conditions. Danke Marlene Böttcher für deine 

Unterstützung und unsere lange Gespräche in KII.  

I would also like to thank my colleagues at the Department of English Linguistics at the 

University of Stuttgart, Nadja Schauffler, Fabian Schubö, and Judith Tonhauser, for their support 

with various things during my PhD time. 

Next, I would like to express my gratitude to all the family members who supported me in this 

venture. My warm спасибо goes to my mother Tatiana and father Andrey Pavlyuchenko as well 

as my grandmother Yulia Minjaeva who is a doctor of science and knows quite a bit about writing 

a dissertation. The next round of gratitude goes to my parents-in-law, Klára Zubánné Szélig and 

László Zubán, as well as to my brother and sister-in-law Rafael Dániel Zubán, Orsolya Zubánné 

Horváth and their daughter Hanna Zubán. I would like to thank my dear husband, Gergely Márton 

Zubán, who supported me all the way from the beginning until the end, and who, by this time, has 

developed some feelings for linguistics, nagyon szépen köszönöm a segítséget és a támogatást! 

Finally, I would like to thank my little cupcake, my daughter Elisa (Элизочка-булочка), for 

teaching me how to manage multiple things simultaneously. 

This thesis was supported by a grant from the German Research Foundation (DFG) to the 

Research Unit “Emerging Grammars” (FOR 2537) in project P7 (project no. 394844953, GZ 

ZE940/2-1; PI: Sabine Zerbian) and (FOR 5137) in project P8 (project number 313607803, ZE 

940/4-1; PIs: Sabine Zerbian, Shanley E. M. Allen, and Oliver Bunk).  

 



 

v 

 Devoted to my daughter Elisa 

 who is growing up to be a 

 heritage speaker of Russian 

 and Hungarian 

 





 

vii 

Contents 

List of Tables xi 

List of Figures xiii 

List of Abbreviations xiv 

Abstract xv 

Deutsche Zusammenfassung xvii 

1 Introduction 1 

2 Heritage Speakers 9 

2.1 Heritage speakers: Definition ............................................................................................ 9 

2.2 Heritage speakers and monolingual speakers: Source of differences .............................. 10 

2.3 Heritage speakers and sociolinguistic factors .................................................................. 15 

2.3.1 Age ............................................................................................................................. 15 

2.3.2 Gender ........................................................................................................................ 18 

2.4 Heritage speakers and different language situations ........................................................ 22 

2.5 Heritage speakers of Russian in the US: History ............................................................. 26 

2.6 Research questions: Summary ......................................................................................... 28 

3 Methodology 29 

3.1 Data elicitation ................................................................................................................. 29 

3.2 Recruitment: Strategies and challenges ........................................................................... 31 

3.3 Data analysis .................................................................................................................... 33 

3.3.1 Transcription and annotation ..................................................................................... 33 

3.3.2 From raw data to statistics ......................................................................................... 34 

3.3.3 General remarks on the statistical analyses of this thesis .......................................... 36 

4 Participants 39 

4.1 Metalinguistic background of the speakers ...................................................................... 39 

4.2 Fluency of Speakers: Speech rate .................................................................................... 45 

4.2.1 Ranges of the average speech rates ............................................................................ 47 

4.2.2 Statistical modeling .................................................................................................... 48 

4.2.3 Results ........................................................................................................................ 49 

4.3 Performance of the heritage speakers .............................................................................. 51 

4.3.1 Nominal inflectional morphology .............................................................................. 51 

4.3.1.1 Case ...................................................................................................................... 51 

4.3.1.2 Gender .................................................................................................................. 55 

4.3.2 Verbs .......................................................................................................................... 58 

4.3.2.1 Analytic verbal constructions .............................................................................. 58 

4.3.3 Aspect ........................................................................................................................ 59 

4.3.4 Clause ......................................................................................................................... 63 

4.3.4.1 Reorganization of complementizes ...................................................................... 63 

4.3.4.2 Possessive reflexive svoj ...................................................................................... 63 



Contents 

viii 

4.3.5 Outside the clause ...................................................................................................... 65 

4.3.6 Phonetics and Phonology ........................................................................................... 66 

5 Corpus study: Global intonation patterns 71 

5.1 Autosegmental-metrical theory of intonation .................................................................. 71 

5.2 Approaches to intonation in monolingual Russian .......................................................... 73 

5.2.1 System of “Intonational Constructions” .................................................................... 73 

5.2.2 Perceptual description of intonation .......................................................................... 74 

5.2.3 Autosegmental-metrical approach ............................................................................. 75 

5.3 Intonation and sociolinguistic factors .............................................................................. 77 

5.3.1 Age ............................................................................................................................. 77 

5.3.2 Gender ........................................................................................................................ 78 

5.4 Intonation and different language situations .................................................................... 80 

5.5 Intonation in heritage languages ...................................................................................... 82 

5.6 Intonation in heritage Russian ......................................................................................... 87 

5.7 Research questions and hypotheses ................................................................................. 95 

5.8 Prosodic annotation ......................................................................................................... 98 

5.8.1 IP-boundaries ............................................................................................................. 99 

5.8.2 Pitch accent placement and type .............................................................................. 100 

5.8.3 Simplification of the data ......................................................................................... 104 

5.9 Results ............................................................................................................................ 106 

5.9.1 Pitch accent type ...................................................................................................... 106 

5.9.1.1 H* ...................................................................................................................... 108 

5.9.1.2 Rising accents .................................................................................................... 110 

5.9.1.3 Falling accents ................................................................................................... 112 

5.9.1.4 L* ....................................................................................................................... 113 

5.9.2 Final boundary tone ................................................................................................. 114 

5.9.3 Pitch accent placement ............................................................................................ 117 

5.10 Summary of the results and discussion .......................................................................... 121 

5.11 Conclusion of this chapter ............................................................................................. 134 

6 Corpus study: Word order choice 137 

6.1 Word order in monolingual Russian: Frequency of different patterns .......................... 137 

6.2 Word order in heritage languages .................................................................................. 138 

6.3 Word order in heritage Russian: Frequency of different patterns ................................. 142 

6.4 Research questions and hypotheses ............................................................................... 149 

6.5 Syntactic annotation ....................................................................................................... 152 

6.5.1 Word order ............................................................................................................... 152 

6.5.2 Clauses ..................................................................................................................... 154 

6.5.3 Further remarks ........................................................................................................ 156 

6.5.4 Simplification of the data ......................................................................................... 160 

6.6 Results ............................................................................................................................ 160 

6.6.1 Descriptive overview ............................................................................................... 160 

6.6.2 Statistical modeling ................................................................................................. 162 



 

ix 

6.6.2.1 SVO.................................................................................................................... 164 

6.6.2.2 OVS.................................................................................................................... 166 

6.6.2.3 SOV.................................................................................................................... 168 

6.7 Summary of the results and discussion .......................................................................... 171 

6.8 Conclusion of this chapter ............................................................................................. 182 

7 Corpus study: Expression of information status 185 

7.1 Definition and primitives of information structure ........................................................ 185 

7.2 Expression of information structure and status in monolingual Russian ....................... 188 

7.2.1 Word order ............................................................................................................... 188 

7.2.2 Intonation ................................................................................................................. 196 

7.3 Expression of information structure and status in heritage languages ........................... 198 

7.4 Expression of information structure and status in heritage Russian .............................. 201 

7.5 Syntactic expression of information status: Research questions and hypotheses .......... 212 

7.6 Annotation of information status ................................................................................... 213 

7.6.1 RefLex scheme ......................................................................................................... 213 

7.6.2 Language-specific principles ................................................................................... 215 

7.6.3 Simplification of the data ......................................................................................... 221 

7.7 Results ............................................................................................................................ 221 

7.7.1 Descriptive overview ............................................................................................... 222 

7.7.2 Statistical modeling .................................................................................................. 225 

7.7.3 New-Given referents: Overall picture ...................................................................... 227 

7.7.3.1 New-given referents: Communicative situations ............................................... 231 

7.8 Summary of the results and discussion .......................................................................... 232 

7.9 Prosodic expression of new subjects: Research questions and hypotheses ................... 239 

7.10 Results ............................................................................................................................ 241 

7.10.1 Intonation of SnewVOgiven utterances ........................................................................ 241 

7.10.1.1 Heritage speakers ............................................................................................... 241 

7.10.1.2 Monolingual speakers ........................................................................................ 244 

7.11 Summary of the results and discussion .......................................................................... 246 

7.12 Conclusion of this chapter ............................................................................................. 248 

8 Overall summary and general outlook 251 

8.1 Short overview of the results ......................................................................................... 251 

8.2 Summary ........................................................................................................................ 253 

8.3 Limitations of the study ................................................................................................. 258 

8.4 Further research ............................................................................................................. 258 

9 Bibliography 261 

Lebenslauf 287 

 

 





 

xi 

List of Tables 

Table 1 Data set for the study: participants .......................................................................................................... 39 

Table 2 Average scores of media use in Russian by HSs and monolinguals ......................................................... 42 

Table 3 Average scores of media use in English by HSs ....................................................................................... 42 

Table 4 Average scores of language skills in Russian by HSs and monolinguals ................................................. 42 

Table 5 Average scores of language skills in English by HSs ............................................................................... 43 

Table 6 Average scores of heritage and majority language use in bilingual families ........................................... 44 

Table 7 Average speech rate (words/minute) across different formalities ............................................................ 47 

Table 8 Attributes of contrasts for the speech rate according to the linear mixed-effect model ........................... 48 

Table 9 Speech rate: outcome of the linear mixed-effect model ............................................................................ 49 

Table 10 Speech rate: Tukey multiple comparison test ......................................................................................... 50 

Table 11 Examples of non-standard sounds by HSs .............................................................................................. 68 

Table 12 Frequent nuclear F0 contours, their AM analysis and semantic interpretation in Russian  ......................  

(compilation by Rathcke, 2009, p. 37) ................................................................................................................... 75 

Table 13 Overall number of pitch accents of different types ............................................................................... 106 

Table 14 Attributes of contrasts for the pitch accent types according to the generalized binominal  .......................  

linear mixed-effect models ................................................................................................................................... 107 

Table 15 Pitch accent H*: Outcome of the generalized binominal linear mixed-effect model ............................ 108 

Table 16 Pitch accent H*: Tukey multiple comparison test (country: gender) ................................................... 109 

Table 17 Rising pitch accents: Outcome of the generalized binominal linear mixed-effect model ..................... 110 

Table 18 Rising pitch accents: Tukey multiple comparison test (country: gender) ............................................. 111 

Table 19 Falling pitch accents: Outcome of the generalized binominal linear mixed-effect model .................... 112 

Table 20 Pitch accent L*: Outcome of the generalized binominal linear mixed-effect model ............................ 113 

Table 21 Overall number of final boundary tones ............................................................................................... 114 

Table 22 High final boundary tone: Outcome of the generalized binominal linear mixed-effect model ............. 115 

Table 23 High final boundary tones: Tukey multiple comparison test (country: formality) ................................ 116 

Table 24 Overall number of words and pitch accents ......................................................................................... 117 

Table 25 Pitch accent placement: Outcome of the generalized binominal linear mixed-effect model ................ 118 

Table 26 Pitch accent placement: Tukey multiple comparison test (country: formality) .................................... 119 

Table 27 Raw and prorated frequencies of WO patterns in the HL, FL3, FL4, and NS sub-corpora  ......................  

(from Kisselev, 2019, p. 160) ............................................................................................................................... 146 

Table 28 Annotation of word order ..................................................................................................................... 152 

Table 29 Annotation of clause type...................................................................................................................... 154 

Table 30 Overall number of occurrences and the distribution of basic word orders across two speaker groups161 

Table 31 Number of occurrences and the distribution of basic word orders across two speaker groups  ................  

in main clauses .................................................................................................................................................... 161 

Table 32 Number of occurrences and the distribution of basic word orders across two speaker groups  ................  

in embedded clauses ............................................................................................................................................ 162 

Table 33 Attributes of contrasts for the word orders according to the generalized binominal  ................................  

linear mixed-effect models ................................................................................................................................... 163 

Table 34 Word order SVO: Outcome of the generalized binominal linear mixed-effect model .......................... 164 

Table 35 Word order SVO: Tukey multiple comparison test ............................................................................... 165 

Table 36 Word order OVS: Outcome of the generalized binominal linear mixed-effect model .......................... 166 



List of Tables 

xii 

Table 37 Word order OVS: Tukey multiple comparison test ............................................................................... 167 

Table 38 Word order SOV: Outcome of the generalized binominal linear mixed-effect model .......................... 168 

Table 39 Word order SOV: Tukey multiple comparison test ............................................................................... 169 

Table 40 Effect of formality and mode on different word orders: Summary ....................................................... 178 

Table 41 Typical organization of referents across different word orders in Standard Russian  ...............................  

(compilation from Kallestinova, 2007, p. 38 and Slioussar, 2007) ..................................................................... 192 

Table 42 Annotation tags of the r-level (Riester & Baumann, 2017, p. 5) .......................................................... 215 

Table 43 Categories annotated in the data regarding the information status of a referent ................................ 219 

Table 44 Number of referents of different types across two speaker groups ....................................................... 222 

Table 45 Number of specific referent combinations across six word orders....................................................... 222 

Table 46 Speaker group: Outcome of the generalized binominal linear mixed-effect model .............................. 226 

Table 47 Number of new-given and other discourse-new referents in different situations ................................. 231 

Table 48 Prosodic patterns of SnewVOgiven and in SgivenVOnew/given by heritage speakers ...................................... 242 

Table 49 Prosodic patterns of SnewVOgiven and in SgivenVOnew/given by monolingual speakers ............................... 244 

Table 50 Comparing the results of the two speaker groups: Global intonation patterns .................................... 251 

Table 51 Comparing the results of the two speaker groups: Word order choice ................................................ 252 

Table 52 Comparing the results of the two speaker groups: Expression of information status .......................... 252 

 



 

xiii 

List of Figures 

Figure 1 Overview of data preparation ................................................................................................................. 35 

Figure 2 The estimates of all main effects and their 95% confidence intervals: Speech rate ............................... 49 

Figure 3 Speech rate: Interaction of country and formality .................................................................................. 50 

Figure 4 Boundaries of intonational phrases: Pitch jump .................................................................................. 100 

Figure 5 Boundaries of intonational phrases: Unit-final lengthening ................................................................ 100 

Figure 6 Different pitch accent types: Example of annotation ............................................................................ 103 

Figure 7 Final boundary tones: Example of annotation ...................................................................................... 104 

Figure 8 The estimates of all main effects and their 95% confidence intervals: H* accent ................................ 108 

Figure 9 Pitch accent H*: Interaction of country and gender ............................................................................ 109 

Figure 10 The estimates of all main effects and their 95% confidence intervals: Rising accents ....................... 110 

Figure 11 Rising pitch accents: Interaction of country and gender .................................................................... 111 

Figure 12 The estimates of all main effects and their 95% confidence intervals: Falling accents ...................... 112 

Figure 13 The estimates of all main effects and their 95% confidence intervals: L* accent ............................... 113 

Figure 14 The estimates of all main effects and their 95% confidence intervals: High final boundary tone ...... 115 

Figure 15 High final boundary tones: Interaction of country and formality ....................................................... 116 

Figure 16 The estimates of all main effects and their 95% confidence intervals: Pitch accent placement ......... 119 

Figure 17 Pitch accent placement: Interaction of country and formality ............................................................ 120 

Figure 18 Pitch track for the annotation of word order ...................................................................................... 159 

Figure 19 The estimates of all main effects and their 95% confidence intervals: SVO word order .................... 165 

Figure 20 Word order SVO: Interaction of country and mode ............................................................................ 166 

Figure 21 The estimates of all main effects and their 95% confidence intervals: OVS word order .................... 167 

Figure 22 Word order OVS: Interaction of country and formality ...................................................................... 168 

Figure 23 The estimates of all main effects and their 95% confidence intervals: SOV word order .................... 169 

Figure 24 Word order SOV: Interaction of country and formality ...................................................................... 170 

Figure 25 Distribution of frequent referent combinations across three word orders .......................................... 223 

Figure 26 The estimates of all main effects and their 95% confidence intervals: Speaker group ....................... 226 

Figure 27 Distribution of new-given referents  (in relation to all clauses with discourse-new referents) .......... 230 

Figure 28 Distribution of new-given referents in four communicative situations  ....................................................  

(in relation to all clauses with discourse-new referents). .................................................................................... 231 

 



 

xiv 

List of Abbreviations 

ACC .............. Accusative case 

AM theory .... Autosegmental-metrical theory of intonation 

CUs ............... Communication units 

DAT .............. Dative case 

DP ................ Determiner phrase 

F ................... Feminine 

F0 ................. Fundamental frequency 

FBT .............. Final boundary tone 

FUT .............. Future 

GEN ............. Genitive case 

HL ................ Heritage language 

HS ................ Heritage speaker 

IKs ................ Intonacionnye konstrukcii (“intonational constructions”) 

INS ............... Instrumental case 

IP .................. Intonation phrase 

IPFV ............ Imperfective 

IPO ............... Institute of perception research 

L1  ................ First language 

L2 ................. Second language 

LILt .............. L2 intonation learning theory 

M .................. Masculine 

MAE_ToBI .. Mainstream American English ToBI 

ML ................ Majority language 

MULTIDIR .. Multidirectional 

NEG ............. Negation 

NOM ............ Nominative case 

NP ................ Noun phrase 

NPA .............. Nuclear pitch accent 

PA ................. Pitch accent 

PFV .............. Perfective  

PL ................. Plural 

POS .............. Part of speech 

PP ................. Prepositional phrase 

PrV ............... Preverbal position of a referent 

PV ................. Postverbal position of a referent 

RUEG ........... Research Unit "Emerging Grammars in Language Contact Situations: A Comparative Approach" 

ToBI ............. Tone and Break Indices 

ToDI ............. Transcription of Dutch intonation 

ToRI ............. Transcription of Russian intonation 

UNIDIR ....... Unidirectional 

YNQ ............. Yes/No question 



 

xv 

Abstract 

The current thesis investigates semi-spontaneous narrations across different formalities 

(formal/informal) and modes (spoken/written) produced by heritage speakers (HSs) of Russian in 

the US and monolingual speakers of Russian. The thesis focuses on global intonational features, 

choice of different word orders, and expression of information status by the two aforementioned 

groups. 

With respect to global intonational features, this thesis showed that heritage and monolingual 

speakers were similar to each other regarding some prosodic patterns (e.g., inventory of pitch 

accents and boundary tones). The two speaker groups also differed from each other regarding 

some other intonational patterns, but this difference was always modulated by other factors such 

as gender and formality (e.g., female HSs produced more H* pitch accents and fewer rising 

accents than monolingual female speakers). Furthermore, the results showed that formality was 

an important factor for some prosodic features but only for monolingual speakers (i.e., final 

boundary tones and pitch accent placement). Formality was not found to have an effect on the 

intonation of HSs. Finally, the results showed a general effect of age and gender that was similar 

for monolingual and HSs regarding some specific intonational patterns (e.g., adolescent speakers 

produced more H* and fewer L* accents than adult speakers; female speakers produced more high 

boundary tones than male speakers).  

With respect to the choice of different word orders, this thesis showed that HSs and 

monolinguals used a similar word order repertoire by producing all six grammatically possible 

word orders. Also, both speaker groups, similar to each other, produced some word orders more 

frequently (SVO, OVS, and SOV) than others (OSV, VSO, and VOS). Further, word orders of 

HSs differed from those of monolinguals only in formal and written situations, but not in informal 

and spoken ones. Specifically, HSs produced more SVO utterances in the written mode and fewer 

OVS utterances in formal situations compared to monolinguals.  

With respect to the expression of information status, it was found that HSs were similar to 

monolingual speakers by producing the three most frequent word orders (SVO, OVS, and SOV) 

according to the underlying discourse principles that govern word order variation in Standard 

Russian (e.g., OVS was frequent with O being discourse-given and S being discourse-new). 

Furthermore, this thesis examined the expression of referents with the order new-given that is 

unexpected in formal monolingual Russian. The results showed that both speaker groups produced 

new-given referents with HSs leading quantitatively. Further, HSs produced new-given referents 

across all investigated formalities and modes. Monolingual speakers produced such referents in 



Abstract 

xvi 

almost every situation except for the formal written one. Finally, the results showed that speakers 

of both groups used speaker-specific prosodic cues to mark new subjects in SnewVOgiven utterances. 

In view of the results of this thesis, I conclude that (1) HSs of Russian in the US are often 

similar to monolingual speakers of Russian regarding some linguistic patterns, (2) when the two 

groups differ from each other, the differences are often found in situations that require formal 

instruction or exposure to a formal register in Russian and (3) linguistic patterns of HSs are often 

similar across different communicative situations, contrary to the patterns of monolingual 

speakers that are frequently influenced by formality and mode. The results in (2) and (3) can be 

possibly explained by the limited exposure of HSs to formal registers in their heritage language. 

The results of this thesis show differences between HSs and monolinguals that relate to the 

frequency of some linguistic patterns and their distribution across different formalities and modes 

rather than qualitative differences between the grammars of the two speaker groups. Thus, HSs 

should be viewed as native speakers of their heritage language who demonstrate variation within 

the native grammar.



 

xvii 

Deutsche Zusammenfassung 

Die vorliegende Doktorarbeit befasst sich mit semi-spontanen Äußerungen von russischen 

Herkunftssprecher:innen (HS) in den USA und von monolingualen Russisch-Sprecher:innen in 

unterschiedlichen kommunikativen Situationen bezüglich Formalität (formell/informell) und 

Modus (gesprochen/geschrieben). Die Arbeit fokussiert sich dabei auf globale Intonation, Wahl 

unterschiedlicher Wortstellungen und Markierung von Informationsstatus durch die beiden 

benannten Sprecher:innengruppen. 

Bezüglich Intonation zeigt diese Arbeit Ähnlichkeiten zwischen HSs und monolingualen 

Sprecher:innen in einigen prosodischen Aspekten (z.B. dem Repertoire von Tonakzenten und 

Grenztönen). Die gefundenen Unterschiede zwischen den Gruppen werden durch andere Faktoren 

z.B. Formalität der kommunikativen Situation und dem Geschlecht der Sprecher:innen moduliert. 

So produzieren weibliche HS mehr monotonal hohe (H*) als steigende Tonakzente im Vergleich 

zu monolingualen Sprecherinnen. Die Formalität der Situation erwies sich als ein wichtiger Faktor 

für einige prosodische Aspekte jedoch nur bei monolingualen Sprecher:innen (Grenztöne und 

Tonakzentplatzierung). Für HS wurde kein Unterschied der Intonation in unterschiedlichen 

Formalitäten beobachtet. Die Ergebnisse zeigen zudem generelle Tendenzen bezüglich des 

Geschlechts und Alters der monolingualen und HSs bezogen auf bestimmte Intonationsmuster 

(z.B. produzieren Jugendliche mehr hohe und weniger tiefe Tonakzente als Erwachsene und 

weibliche Sprecherinnen produzieren mehr hohe Grenztöne als männliche). 

Bezüglich der Wortstellung zeigt die Arbeit ein vergleichbares Repertoire von sechs 

grammatikalisch möglichen Wortstellungen im Russischen für monolinguale und HSs. Dabei 

zeigt sich das einige Wortstellungen häufiger auftreten (SVO, OVS, SOV) als andere (OSV, VSO, 

VOS). Unterschiede zwischen den Sprechergruppen finden sich vor allem in der formellen 

Situation und im geschriebenen Modus und nicht im Informellen und Gesprochenen. Insbesondere 

die SVO Struktur wurde von HS häufiger in geschriebenen Äußerungen und die OVS Struktur 

seltener in formellen Äußerungen im Vergleich zu monolingualen Sprecher:innen produziert. 

Bezüglich des Informationsstatus zeigt sich eine häufige Verwendung derjenigen 

Wortstellungen SVO, OVS und SOV, die den Prinzipien des Diskurses im Standardrussischen 

entsprechen (z.B. OVS war häufig in Kontexten in denen gegebene O den neuen S vorausgehen) 

sowohl von monolingualen als auch von HS. Diese Arbeit zeigt auch unerwartete Strukturen, in 

denen die neuen Referenten den gegebenen voraus gehen. Diese Struktur wird überwiegend von 

HS produziert, findet sich aber auch in den Äußerungen von monolingualen Sprecher:innen mit 



Deutsche Zusammenfassung 

xviii 

Ausnahme der formell geschriebenen Sprache. Eine weiterführende Untersuchung der Prosodie 

zeigt individuelle intonatorische Markierung neuer Subjekte in der Struktur SneuVOgegeben. 

Die Ergebnisse dieser Arbeit lassen die folgenden Schlüsse zu: (1) HS des Russischen in den 

USA ähneln monolingualen Sprecher:innen in einigen linguistischen Bereichen. (2) Unterschiede 

zwischen den Gruppen finden sich in den kommunikativen Situationen und Modi, die formelle 

Instruktionen oder dergleichen in der Russischen Sprache voraussetzen. (3) Sprachliche 

Strukturen in HS gleichen sich häufig über Situationen und Modi, während monolinguale 

Sprecher:innen Sensibilität für Formalität und Modus zeigen. Die Ergebnisse zu (2) und (3) lassen 

sich durch eine begrenzte oder gänzlich fehlende Anwendung eines formellen Registers im 

herkunftssprachlichen Kontext erklären. 

Die Ergebnisse dieser Arbeit zeigen Unterschiede zwischen HS und monolingualen 

Sprecher:innen in der Häufigkeit bestimmter linguistischer Phänomene und deren Verteilung im 

Zusammenspiel mit Formalität und Modus. Sie lassen jedoch auf keine qualitativen Unterschiede 

zweier Sprachsysteme schließen. Die Arbeit liefert damit weitere Hinweise für den Status von HS 

als Muttersprachler ihrer Herkunftssprache die spezifische Variation aufzeigt. 



 

1 

1 Introduction 

Heritage speakers (HSs) form a unique group of speakers due to the early acquisition of one 

language at home (their HL) and the subsequent or simultaneous acquisition of another language 

of the larger society (the majority language, ML) which typically becomes dominant.  

HSs form quite a heterogeneous group of speakers regarding their proficiency in their HL: 

some HSs may be very fluent in their HL while some HSs may only comprehend their HL but not 

speak it (Polinsky, 2015, p. 10). HSs typically receive the main input in their HL from their family 

members in an informal spoken form (Polinsky, 2015; Flores & Rinke, 2020). 

Heritage language bilingualism has received a great deal of attention over the past two decades 

for at least two reasons. First, it is highly relevant in our current society. The number of HSs has 

risen substantially in recent years due to increased global immigration. Thus, the language needs 

of HSs are increasingly salient in the education system. A main focus is early support for 

developing linguistic skills in the ML while support for maintaining the HL is also increasing in 

many school systems. Second, the study of HSs is particularly relevant linguistically since it 

presents excellent testing grounds for the theories of language contact, acquisition and 

maintenance and language processing theories, as well as for formal linguistics. 

The current thesis analyzes semi-spontaneous productions across different communicative 

situations (formal/informal, spoken/written) in the two speaker populations, HSs of Russian in the 

US and monolingual speakers of Russian. This thesis investigates three linguistic areas, namely 

word order, information status, and global intonation patterns. 

HSs were often found to show more differences from monolingual speakers in areas such as 

morphology or syntax than in the production of speech sounds, leading to what was called a 

“phonetic advantage” (Polinsky, 2018, pp. 115-116; Chang, 2021, pp. 582-3). Such an advantage 

may arise as a result of early exposure to a HL (Kan, 2021, p. 74). However, early exposure to an 

HL does not guarantee a monolingual-like production of segmental and suprasegmental aspects 

during the lifetime of HSs since speech sounds as well as stress or prosodic differences might be 

subject to variation and change (Polinsky, 2018, p. 116; Kan, 2021, p. 74). 

Impressionistically, HSs are typically perceived somewhere between monolingual speakers 

and L2 learners, forming a separate speaker group. The speech of HSs is often rated as more 

accented than that of monolingual speakers and less accented than that of L2 learners1, but what 

exactly distinguishes HSs from the monolingual group remains unclear (Flores & Rato 2016; 

Kupisch et al., 2014; Polinsky, 2018; Chang & Yao, 2016).  

                                                 
1 An L2 learner is considered to be a learner of their second language in adulthood (Polinsky, 2018, p. 114).  



Chapter 1. Introduction 

2 

In addition to the differences in the segmental domain, studies on suprasegmental aspects such 

as stress, rhythm, and intonation report that HSs differ from both monolingually-raised speakers 

and L2 learners (for a detailed overview see Chang, 2021, pp. 593-597). In first language 

acquisition, intonation was found to be fully acquired late. Thus, for a child who grows up being 

exposed to several languages simultaneously at a young age, there is considerable room for cross-

linguistic interactions and variation in both the HL and the ML. The cross-linguistic influence 

might arise regarding the inventory and distribution of categorical phonological elements 

(systemic dimension), their functionality (semantic dimension), phonetic implementation 

(realisational dimension), and frequency of use (frequency dimension) (Mennen, 2015). HSs were 

found to differ from monolingually-raised speakers in every dimension of intonation, and the 

differences were attributed to the influence from the ML or to some general effects of bilingualism 

related to processing, fluency or decreased language exposure or use (see Colantoni et al., 2016; 

Kim & Repiso-Puigdelliura, 2021; Kim, 2023 & Kachkovskaia et al., 2022 on the influence of 

MLs on a HL; see Zuban et al., 2023 or Zerbian et al., 2022 on general effects of bilingualism). 

It has to be acknowledged that intonation of HLs in general and heritage Russian in particular 

has received far less attention than other areas of HL grammar (Rao & Polinsky, 2024). Further, 

the majority of the studies focus on one particular phenomenon such as question intonation, Uptalk 

or intonation of different information-structural categories (e.g., Dehé, 2018; Dehé & Rommel, 

2024; Zuban et al., 2023; Kim & Repiso-Puigdelliura, 2021; Kim, 2023; Kim, 2019; Zerbian et 

al., 2022; Calhoun et al., 2024; Feldhausen & del Mar Vanrell, 2024). However, it is less clear 

what global intonation features of HSs are, i.e., the overall inventory of pitch accents and boundary 

tones as well as their type and distribution. Thus, being one of the first studies on global 

intonational patterns of HSs of Russian, this thesis aims to shed more light on the possible reasons 

behind the perceived accent of HSs and why HSs are often seen as a group of speakers that fall 

between monolingual and L2 speakers (Rao & Polinsky, 2024). 

Furthermore, age and gender-related differences have been extensively reported by the studies 

on prosody in monolingual populations, and age and gender are among the crucial factors that 

influence language variation (Tagliamonte, 2016; Nagy, 2017). The inclusion of speakers of 

different ages and genders will allow us to see whether the differences between HSs and 

monolinguals are modulated by these sociolinguistic factors.  

Finally, intonation was found to differ not only between speakers of different genders and ages 

but also between different speaking styles (e.g., formal vs. informal). Research on the influence 

of different speaking styles on intonation of HSs is extremely scarce. A study by Comstock (2018) 

on one HS of Russian found that the HS did not vary their linguistic patterns in different interviews 



 

3 

while Zerbian et al. (2024) report mixed results with respect to register differentiation across HSs 

of Russian in the US and monolinguals. On the one hand, HSs were sensitive to the differences 

between formal and informal registers in a way similar to monolingual speakers regarding some 

prosodic patterns (e.g., both speaker groups produced more pitch accents on the verbs in informal 

situations). On the other hand, in the case of the size of intonational phrases, HSs did not show 

any effect of formality, contrary to monolingual speakers whose intonational phrases were shorter 

in formal situations. Taking the scarcity of previous studies on the possible effect of different 

speaking styles on intonation and their mixed results, this thesis will contribute to the field of 

heritage grammar by including speech of different formalities. 

This thesis will investigate general intonation patterns of adolescent and adult HSs of Russian 

in the US as well as monolingual speakers of Russian in semi-spontaneous formal and informal 

narrations. The present thesis is, to my knowledge, one of the first detailed investigations of the 

global prosodic features of HSs of Russian based on a large data set and including multiple factors 

that may influence speakers’ choice of a particular intonation pattern (i.e., bilingualism, formality, 

age and gender). This thesis will examine pitch accent type (H*, L*, rising and falling accents), 

boundary tone (H% or L%), and pitch accent placement produced by HSs of Russian in the US. 

Although both monolingual Russian and English are intonational languages, they are not 

identical regarding the specific intonational patterns. For instance, it has been reported that 

monotonal pitch accents H* and L* are frequently used in monolingual English while bitonal 

accents are typical for monolingual Russian (Comstock, 2018). Such differences between 

monolingual English and Russian may give rise to differences between HSs and monolingual 

speakers, if there are traces of cross-linguistic influence from the ML. 

Moving to the syntactic properties of HLs, word order was reported to be susceptible to change 

and transfer from the ML (e.g., Heine, 2008, p. 34; Polinsky, 2018, p. 273). Further, languages 

that allow word order variation show less varied order patterns in HLs and increased use of basic 

word orders (cf. Polinsky, 2018, p. 273; Montrul, 2015, p. 82; Lohndal, 2021). In languages that 

allow for word order alternations, such alternations are typically highly governed by information 

structure/status. Studies that examined the encoding of information structure through word order 

in HLs report different findings. Some researchers report on the substantial differences between 

HSs and monolinguals (e.g., Zapata et al., 2005; Laleko & Dubinina, 2018) while other studies 

found only subtle differences between the former and the latter group or no differences at all (e.g., 

Hoot & Leal, 2023; Sequeros-Valle et al., 2020; Ionin et al., 2020). 

Monolingual Russian has a basic SVO order in a broad focus context, but unlike monolingual 

English, allows for remarkable word order alternations. Word order variation depends on many 



Chapter 1. Introduction 

4 

factors, among which information structure/status and intonation are claimed to be the most 

influential (Kallestinova, 2007; Slioussar, 2007). Furthermore, monolingual Russian follows a 

cross-linguistic tendency to place given referents before new ones in formal speech. In informal 

speech, the given-new order can be violated since intonation plays a stronger role in the expression 

of information structure/status (Sirotinina, 2003, p. 152). Specifically, the information status of a 

referent, similar to English, can be expressed solely by intonation without any word order 

permutations (Kallestinova, 2007). Furthermore, different communicative situations as well as 

clause type were reported to be important for the use of particular word orders (Sirotinina, 2003; 

Slioussar, 2007; Bailyn, 2012). 

Numerous studies on heritage Russian reported that HSs of Russian in the US increased the 

use of SVO and decreased the use of non-SVO word orders (e.g., Polinsky, 1995, 2006; Isurin & 

Ivanova-Sullivan, 2008; Laleko & Dubinina, 2018). However, some recent studies showed that 

HSs did not differ from monolinguals across the board, but the more frequent use of the SVO was 

modulated by other factors such as clause type, i.e., HSs were found to be similar to monolinguals 

in main clauses but not in embedded ones, where they predominately produced SVO orders 

(Zuban et al., 2021). Moreover, some studies such as Martynova et al. (in press) reported that HSs 

of Russian in the US and monolingual speakers did not differ from each other regarding the choice 

of the OV/VO orders.  

Although word order in monolingual Russian is discourse-driven, fewer studies focused on the 

syntactic expression of information structure/status by HSs of Russian. Some studies make 

observations that HSs do to not always produce contextually appropriate word orders (Brehmer 

& Usanova, 2015 on heritage Russian in Germany; Zuban et al., 2021; Martynova et al., in press). 

Some studies report a substantial amount of contextually infelicitous word orders produced by 

HSs, up to 20% or 30% of all the word orders (Laleko & Dubinina, 2018). Finally, contrary to the 

aforementioned studies, there is also evidence that HSs of Russian generally know the discourse 

principles that govern word order variation in Russian (production study by Kisselev, 2019; 

perception study by Ionin et al., 2020). 

Previous findings on the frequencies of different word order patterns as well as their use to 

express information structure/status generally show that one can expect that HSs and monolingual 

speakers of Russian would differ from each other. Furthermore, the two speaker groups are 

expected to differ from each other following the predictions of the Interface Hypothesis according 

to which phenomena at an external interface (here syntax and discourse) are predicted to have 

increased variability under language contact (Sorace, 2011). 



 

5 

It is important to point out that the studies on word order and information structure/status in 

heritage Russian did not usually look at the multiple factors that could influence the word order 

choice and the expression of information structure/status by HSs (e.g., clause type, different 

formality or mode) but typically only concentrated on the differences between the speaker groups 

(i.e., HSs vs. monolingual speakers) (e.g., Isurin & Ivanova-Sullivan, 2008; Laleko & Dubinina, 

2018; Kisselev, 2019). However, recent studies show that other factors such as clause type or 

object role can influence the word order choice of HSs (see Zuban et al., 2021 and Martynova et 

al., in press). 

Studies regarding HSs’ knowledge of different registers provide different results. On the one 

hand, HSs were often found to not vary their linguistic patterns in different language situations 

(e.g., Schroeder et al., in press on postverbal elements in heritage Turkish; Alexiadou et al., 2022 

plural indefinite article kati). However, there is also evidence that HSs have an underlying 

knowledge of different registers (Pashkova et al., 2022; Martynova et al., 2024, both on different 

clause types, the former is on heritage German while the latter is on heritage Russian). As in the 

case of intonation of HSs, this thesis will contribute to research on the possible effect of different 

language situations on word order and information status of HSs of Russian. 

Since information structure/status is not only influenced by word order in monolingual Russian 

but also by intonation, the role of intonation in the expression of information structure by HSs of 

Russian in the US is an interesting topic. To my knowledge, prosodic expression of information 

structure/status in HSs of Russian has received almost no attention in current studies except a 

master’s thesis by Ivashchyk (2019) who investigated a small number of HSs. Investigation of 

intonation will give a better picture of the structure of the HL and the strategies of HSs in their 

HL, e.g., if the information structure/status is not always expressed by the word order, it is 

important to determine whether other strategies such as intonation are at play. 

Finally, some of the factors that could influence the word order choice and the expression of 

information structure/status by HSs of Russian in the US, to my knowledge, have not been looked 

at (i.e., different age groups). 

Against this background, this thesis investigates the word order choice by HSs of Russian in 

the US as well as monolingual speakers of Russian considering the influence of bilingualism, 

formality, mode, clause type, and age. Furthermore, this thesis will examine how HSs and 

monolingual speakers use the three most frequent word orders (SVO, OVS, SOV) to express 

information status across different formalities, modes, and age groups. 

Further, this thesis will concentrate on the combinations of referents that are not expected to 

appear in formal Standard Russian, namely new-before-given. I will examine whether such 



Chapter 1. Introduction 

6 

unexpected instances appear in the data, and if they do, where it typically happens (i.e., with which 

word orders, in which formality and mode). Moreover, I will look at the intonation of the 

unexpected new referents.  

It is important to acknowledge that HSs of different languages were often found to differ from 

monolingual speakers regarding various linguistic phenomena, and these differences were often 

discussed as HSs having incomplete acquisition of their HL or its attrition. Also, if HSs differed 

from monolingual speakers such differences were often classified as errors, and the HL was often 

viewed as vulnerable (see Polinsky & Scontras, 2019; Polinsky, 2018; Montrul, 2008, 2015 for 

discussion). 

The recent studies on bilingual speakers, however, view HSs as native speakers of their HL 

who show variation within native grammars. Under this view, the differences that arise between 

HSs and monolingual speakers are not errors but innovations, and their HL is not vulnerable but 

dynamic (e.g., Embick et al., 2020; Flores & Rinke, 2020; Rothman & Treffers-Daller, 2014; 

Kupisch & Rothman, 2018; Wiese et al., 2022). In line with the current studies, this thesis seeks 

to promote a positive perspective on bilingualism by pointing out that HSs are native speakers of 

their HL, and they should not be viewed differently only because they do not always converge 

with monolingual speakers. 

To sum up, the current thesis aims at gaining new insights into HSs’ grammar and promote a 

positive view on bilingualism investigating different linguistic domains in a large-scale study that 

includes multiple factors that can influence the performance of HSs.  

This thesis will contribute to the theories of language contact (L2 Intonation Learning Theory, 

Mennen, 2015; Interface Hypothesis, Sorace, 2011) and will show how the HL is maintained 

among HSs of Russian in the US in the three investigated areas. For the benefit of theoretical 

linguistics, the obtained results may help identify stable and dynamic domains of Russian HL and, 

thus, contribute to the ongoing research of the HSs’ linguistic system. Furthermore, if HSs and 

monolinguals show differences, this thesis will try to shed light on the source behind these 

differences. 

Finally, the results of this thesis can be used in HL education to improve the existing programs 

teaching HLs. The topics of my research might be particularly challenging for a HL classroom 

since they either require the knowledge of different contexts (word order variation for information 

structural purposes) or they require the knowledge of subtle nuances for their mastery (i.e., 

intonation patterns). 



 

7 

The structure of this thesis 

Chapter 2 starts with a definition of heritage speakers and further presents an overview of 

factors that may influence their linguistic behavior (age, gender, different communicative 

situations). It also gives a short description of the history of heritage speakers of Russian in the 

US due to the relevance of these speakers for the current thesis. Finally, chapter 2 finishes with a 

short overview of the research questions this thesis will answer.  

Chapter 3 addresses the general methodology used in this thesis for data collection as well as 

data analysis. Chapter 4 focuses on participants of this thesis regarding their metalinguistic 

background, fluency, and general performance. Chapters 5, 6, and 7 present three corpus studies 

on various linguistic topics. Chapter 5 examines global prosodic patterns of heritage and 

monolingual speakers of Russian. Chapter 6 investigates the word order choice of the two 

aforementioned speaker groups while Chapter 7 deals with the expression of information status. 

Chapter 8 provides an overall summary of the main findings of this thesis, points to some 

limitations of this thesis and finishes with further possible directions. 





 

9 

2 Heritage Speakers 

This chapter is devoted to the speaker group that is at the core of this thesis, namely heritage 

speakers. It starts with the definition of HSs and moves to a brief summary of studies that report 

on the differences and similarities between heritage and monolingual speakers in various linguistic 

areas. Furthermore, the chapter presents the possible reasons for the differences between HSs and 

monolingual speakers. Moreover, this chapter motivates the research questions of this thesis by 

describing how age and gender, being the main social categories crucial for language variation 

and change, may influence the way HLs are used. Next, this chapter describes the performance of 

HSs regarding different formalities (formal vs. informal) and modes (spoken vs. written). 

Furthermore, this chapter gives a short historical overview of HSs of Russian in the US since these 

speakers are of direct relevance to this thesis. Finally, this chapter presents a general summary of 

the research questions that will be addressed in this thesis. 

2.1 Heritage speakers: Definition 

Definitions of HSs vary depending on the perspective. HSs can be broadly identified based on 

their cultural heritage rather than a HL. In this case, HS may not speak their HL at all (Van Deusen-

School, 2003). However, such a definition is not of interest in this dissertation since only HSs who 

possess linguistic knowledge of their HL will be looked at. 

According to a definition that takes the knowledge of a HL into account, “HSs are individuals 

who were raised in homes where a language other than the dominant community was spoken and 

thus possess some degree of bilingualism in the heritage language and the dominant language” 

(Polinsky, 2015, p. 8). 

HSs are also referred to as early bilinguals (Zárate-Sández, 2015), simultaneous or sequential 

bilinguals (Montrul, 2012), unbalanced bilinguals (Baker & Jones, 1998) or incomplete acquires 

(Montrul, 2002). Early bilinguals learn their home language in childhood or since birth (early); 

HSs who learn both their HL and the ML since birth are called simultaneous bilinguals; HSs may 

start learning their HL before their ML, and such HSs are usually referred to as sequential 

bilinguals. HSs are also called unbalanced in the sense that they may be much more proficient in 

their ML compared to their HL; HSs’ knowledge of their HL is sometimes referred to as 

“incomplete” pointing out the fact that HSs may have never fully learned their HL in comparison 

with their parents or monolingual speakers of their HL. 

I will adopt a definition of a HL by Rothman (2009, p. 156) according to whom a HL “is a 

language spoken at home or otherwise readily available to young children, and crucially this 

language is not a dominant language of the larger (national) society.” Also, a person can be called 



Chapter 2. Heritage Speakers 

10 

a HS only if they have some knowledge of their HL that was acquired naturalistically. Besides, 

the competence of a HS in their HL is expected to differ from a monolingual speaker of a similar 

age (Rothman, 2009, p. 156). HSs typically have some knowledge of their HL, but they are usually 

more dominant in their ML (Polinsky & Scontras, 2019; Montrul, 2012). HSs form quite a 

heterogeneous group of speakers regarding their proficiency in HL from those who can barely be 

distinguished from monolingual speakers to those who are hardly competent in their HL (Polinsky, 

2015, p. 10).  

HLs are usually not supported outside the family in the way MLs are. HSs typically receive 

their education in their ML but not in their HL. If HSs do not receive any formal support in their 

HL, they are usually not literate in their HL (Rothman, 2009, p. 157; Polinsky, 2015). Those HSs 

who are literate in their HL are predicted to be more fluent in reading than writing (Polinsky, 

2015).  

The input and the use of a HL may dramatically decrease since HSs feel pressure to assimilate 

into the surrounding dominant society and begin to speak the ML at home more frequently. The 

decrease in HL input may affect the language command of HSs compared to monolingual speakers 

of a similar age (Rothman, 2009, p. 157). Further, HSs’ input is usually restricted in terms of 

language use and number of speakers (mainly everyday informal topics, and small groups of 

family members that do not show the whole language range). Furthermore, HSs may be exposed 

exclusively to one language variety that differs from the standard one (Polinsky, 2015; Flores & 

Rinke, 2020). 

2.2 Heritage speakers and monolingual speakers: Source of differences 

By the time HSs become adolescent or early adult speakers their HL usually differs from the 

language of monolingual speakers. Generally, HSs have good speaking and listening abilities and 

a large, but quite specific, vocabulary, and they are fluent overall in their HL. The language of 

HSs is characterized as informal2 and stereotypical (i.e., it contains a set of words or expressions 

that usually appear together) (Rothman, 2009, p. 157).  

Among the language areas in which HSs were reported to be particularly different from 

monolingual speakers are nominal and verbal morphology (Montrul, 2015, p. 54). In terms of 

nominal morphology, HSs were frequently found to inconsistently mark gender, number and case 

in their HLs, produce regular plural forms instead of irregular ones, omit overt case marking, and 

use fewer cases than in the baseline language. In terms of verbal morphology, HSs were found to 

                                                 
2 The terms “informal”, “colloquial” and “vernacular” are used interchangeably in this thesis referring to language 

that people use in everyday situations (e.g., at home) as opposed to language used in official contexts (e.g., in court). 



 

11 

differ from monolinguals regarding aspect and mood, but their use of tense and agreement was 

reported to be more monolingual-like (see Zombolou, 2011; Montrul & Potowski, 2007; Albirini 

et al., 2013; Polinsky, 2006, 2008a, b, 2018 pp. 177-179; Bolonyai, 2007; Montrul et al., 2012; 

Laleko, 2010; Montrul, 2002, 2007, 2009). 

In the research on HLs, it is common to compare HSs against a monolingual standard, and if 

HSs differ from monolinguals regarding a particular phenomenon, such differences are often 

classified as errors, deviances, or mistakes. If speakers of the ML show some changes, but HSs 

do not, then HSs are described as conservative (Polinsky, 2018, p. 129). Different outcomes 

between HSs and monolinguals are typically attributed to the quantity and the quality of the input 

or to the economy of online resources which can result in the avoidance of ambiguity, resistance 

to irregularity, and the shrinking of structure (Polinsky & Scontras, 2019, pp. 10-11). Language 

areas where such different outcomes occur are often described as vulnerable. The potential sources 

for such differences that are typically discussed in the literature are the following: transfer, 

attrition, and divergent attainment (Polinsky & Scontras, 2019, p. 2; Polinsky, 2018; Montrul, 

2008, 2015). 

Transfer from the ML was reported for different language domains of HLs (Polinsky, 2018, 

chapter 2; Brehmer & Usanova, 2015; Montrul, 2010). However, transfer was not always found 

to be the reason for differences between HSs and monolinguals (e.g., Zuban et al., 2023; Zuban et 

al., 2021; Polinsky, 2011).  

Another possible reason for the differences between HSs and monolinguals is attrition: HSs 

may have acquired a particular language phenomenon but then lost some or all the features of this 

phenomenon (Polinsky & Scontras, 2019, p. 3).  

Finally, it is also possible that the system acquired by HSs is not similar to the one acquired by 

monolingual speakers, and in this case, one speaks of divergent attainment (or incomplete 

acquisition) (Polinsky & Scontras 2019, p. 4). According to Polinsky & Scontras (2019), it is often 

not easy to determine whether transfer from the ML or a divergent attainment is at play. If HSs 

don’t produce a phenomenon in their HL that does not exist in their ML, but exists in the baseline 

it is not clear whether it is a result of transfer or a divergent attainment. 

Other recent approaches promote a positive, multilingual perspective on HSs by reanalyzing 

the concept of vulnerability of linguistic domains as linguistic domains that are dynamic. In line 

with this approach, HSs do not suffer from language attrition or divergent attainment but create 

innovative structures. Under this view, HSs show variation within native grammars (Embick et 

al., 2020; Flores & Rinke, 2020; Rothman & Treffers-Daller, 2014; Guijarro-Fuentes & Schmitz, 

2015; Kupisch & Rothman, 2018; Pascual y Cabo & Rothman, 2012; Flores, 2015; Rinke & 



Chapter 2. Heritage Speakers 

12 

Flores, 2014; Wiese et al., 2022). Rothman and Treffers-Daller (2014) and Kupisch and Rothman 

(2018) point out that HSs are native speakers of their HLs, and they should not be viewed 

differently only because they do not always converge with the monolingual standard to which they 

are often compared.  

In a paper by Pascual y Cabo and Rothman (2012), the authors challenge the concept of 

incomplete acquisition that has been used to describe the competence of HSs for quite a long time. 

First, it is impossible to find out whether HSs that were tested as adults did not acquire a particular 

phenomenon or whether they acquired it but later forgot it (i.e., attrition). Second, even if someone 

were to show (in a longitudinal study) that HSs acquired a particular phenomenon during their 

childhood but then “lost” it later on, such results would not necessarily need to be explained in 

terms of incomplete acquisition. Instead, one could say that HSs follow a different path of 

complete development (Pascual y Cabo & Rothman, 2012, p. 453). Pascual y Cabo and Rothman 

(2012) suggest that the differences in language use shown by HSs do not have to be considered as 

deficits but should be viewed as outcomes in their own right (see Pires, 2011; Kupisch & Rothman, 

2018 for a similar idea).  

Furthermore, differences between monolingual and HSs may be related to the type of speech, 

namely colloquial vs. formal. As noted earlier, HSs are typically exposed to colloquial variants of 

a language and usually do not receive formal education in their HL (Flores & Rinke, 2020; 

Lohndal et al., 2019; Pascual y Cabo & Rothman, 2012; Flores, 2015). As a result, differences 

that are found between HSs and monolinguals may be attributed to the differences in language 

situations, i.e., HSs are more likely to differ from monolinguals in formal situations than in 

informal ones. Thus, it is important that HSs’ language use is not only compared to the standard 

language but also to its vernacular variant (see Flores & Rinke, 2020, p. 25 for a similar idea). 

A study by Rothman (2007) supports the abovementioned point. Rothman (2007) examined 

the knowledge of inflected infinitives by 11 HSs of Brazilian Portuguese with the ML American 

English who did not receive a consistent formal education in Portuguese. Inflected infinitives were 

argued not to be a property of colloquial Portuguese, but they exist in formal standard Brazilian 

Portuguese. The results of the study showed that HSs (most of the tested participants) did not have 

a knowledge of inflected infinitives. Rothman (2007) explains such results not by attrition or 

incomplete acquisition, but by the fact that HSs were not exposed to the formal variety of 

Portuguese, and in this case, they simply did not acquire inflected infinitives at all.  

Similarly, Kupisch and Rothman (2018) found that differences that arose between HSs and 

monolinguals were not caused by the incomplete acquisition of the HLs, but by other factors such 

as quality of input and general inaccessibility to formal education in a HL. Kupisch and Rothman 



 

13 

(2018) summarized a series of studies that used the same methodology to investigate four 

linguistic phenomena (gender marking, adjective placement, determiner in generic expressions, 

and voice onset time) by HSs of French and Italian residing in Germany. Importantly, HSs of 

French attended a Francophone school in Germany that followed the curriculum of the French 

Ministry of Education and allowed the students to graduate from two schools simultaneously 

(German Abitur and French Baccalauréat). HSs of Italian living in Germany attended monolingual 

German schools and only some of the HSs attended Sunday schools during their childhood where 

they learned how to read and write in Italian.  

The results of these studies revealed that HSs of French were closer to monolingual speakers 

than HSs of Italian (Kupisch & Rothman, 2018, pp. 570-573). Such results are explained with 

reference to formal education. Texts that are discussed in schools go beyond colloquial language 

(e.g., classic literature or technical texts) and thus learners are exposed to various grammatical 

structures and vocabulary. Moreover, teachers at bilingual schools are more likely to speak a 

standard variety of a language that differs from the language HSs hear at home (Kupisch & 

Rothman, 2018, p. 576). HSs of French attended Francophone schools for more than ten years and 

were familiar with academic (formal) language and had experience taking exams in French 

(Kupisch & Rothman, 2018, p. 573). HSs of Italian did not have the same linguistic experience as 

HSs of French, and, consequently, the former speakers were more distinct from monolingual 

speakers than the latter. Kupisch and Rothman (2018) conclude that consistent education gives 

HSs quantitatively and qualitatively more opportunities to converge more closely with 

monolingual speakers. 

Wiese et al. (2022) consider HSs to be native speakers of both HL and ML with evidence from 

a large-scale cross-linguistic study. The study examines the productions of HSs with heritage 

Greek, Russian, and Turkish and majority American English and German as well as heritage 

German in the US and monolingual speakers from Greece, Russia, Turkey, the United States, and 

Germany. The datasets include speakers’ productions in different language formalities and modes: 

formal and informal as well as written and spoken. 

Wiese et al. (2022) report three main results. First, not only HSs, but also monolingual 

speakers, were found to produce non-canonical phenomena in the areas of intonation (final 

boundary tones in yes-no questions in Russian), pragmatics (referents introduction in English and 

position of new referents in Turkish) and, morphology and syntax (word order and bare NPs in 

German, formation and use of participles in Russian). 

Second, monolingual speakers sometimes showed a higher degree of lexical and 

morphosyntactic inter-speaker variation compared to HSs. 



Chapter 2. Heritage Speakers 

14 

Third, in MLs, non-canonical patterns typically appeared in informal or/and spoken situations, 

and this trend held for HSs and monolingual speakers (e.g., non-canonical new and given referents 

in majority English were associated with the spoken mode rather than with the written mode across 

bilinguals and monolinguals; non-canonical V3 patterns in majority German were associated with 

informal situations rather than with formal ones). Some non-canonical patterns in informal 

situations were produced more frequently by bilingual speakers compared to monolinguals. 

As for HL use, Wiese et al. (2022) report that some patterns typical for informal situations in 

the narrations of monolingual speakers were extended to formal situations in the narrations of HSs 

leading to register-leveling processes (e.g., HSs of German in the US produced non-canonical V3 

word order patterns or bare NPs in both formal and informal situations roughly equally frequently 

while monolingual and bilingual speakers in Germany produced such patterns in informal 

situations but not in formal situations). Register-leveling is presumably caused by the lack of 

formal education in the HLs. 

Wiese et al. (2022) conclude that HSs and monolinguals do not have distinct grammatical 

patterns, but rather differ from each other quantitatively and have different register distributions. 

Thus, HSs should be viewed as native speakers of both HL and ML. Moreover, the study revealed 

that monolingual speakers produced patterns that were reported to be non-canonical according to 

the literature, and these patterns typically occurred in informal and spoken situations. Therefore, 

it is important to include the data of monolingual speakers as well as informal and spoken 

situations for the investigation of HSs’ productions since it may turn out that HSs show no effect 

of bilingualism in these situations and behave just like monolingual speakers. 

Furthermore, although language variation was reported to be a feature of all language 

communities it was found that colloquial language serves as the main source of linguistic variation 

whereas formal and written language is less prone to variation. Since HSs are usually not exposed 

to the formal variety of their HL, they tend to increase variation that already exists in monolingual 

speech (Flores, 2015, p. 254). 

Another important issue to tackle is to whom should adult HSs be compared. Current studies 

emphasize that HSs should not be directly compared to monolingual homeland speakers since 

monolinguals do not serve as their input. HSs should rather be compared to first-generation 

immigrants since this group of speakers provides the input for HSs in their HL. Besides, the 

language spoken by first-generation immigrants may differ from the language of monolingual 

homeland speakers (Polinsky, 2018, pp. 10-13; Polinsky & Scontras, 2019; Benmamoun et al., 

2013a, b; Madsen, 2018, the last three studies are cited by Polinsky & Scontras, 2019, p. 4). 



 

15 

However, if one investigates acquisition of a HL a comparison with homeland children is 

appropriate (Polinsky, 2018, pp. 11-12). 

The idea that HSs should not be measured against the monolingual standard does not mean that 

monolingual data should not be considered at all. Data from monolingual speakers are also fruitful 

since monolinguals show language variation as well, and one can compare degrees of variation in 

monolingual and heritage groups without judging how “complete” the grammar of HSs is 

(Guijarro-Fuentes & Schmitz, 2015, p. 244). Such variation can be particularly detected if not 

only formal, but also informal language situations are included in the analysis. Thus, the inclusion 

of the monolingual group and different language situations (formal and informal) will allow for 

distinguishing contact-induced phenomena from general patterns of variation. 

To sum up, a bulk of studies name transfer, attrition, and incomplete acquisition to be the 

reasons for the differences between HSs and monolingual speakers. According to more recent 

approaches HSs do not show language attrition or incomplete acquisition but create innovative 

structures. Furthermore, current studies emphasize the importance of the vernacular language in 

HL research since HSs are typically not exposed to the formal language. This thesis adopts a 

positive approach to bilingualism and views HSs as the native speakers of their HL who show 

variation within native grammars. 

2.3 Heritage speakers and sociolinguistic factors 

Since gender and age have been claimed to be among the most crucial factors that influence 

language variation these sociolinguistic factors will be discussed in more detail for bilingual 

populations. Furthermore, gender-related differences have been reported to be particularly 

prominent in the area of phonetics, phonology, and intonation. Therefore, the subchapter on 

gender will mainly concentrate on the studies related to the abovementioned linguistic areas.  

2.3.1 Age 

Age was reported to be one of the crucial factors that influences language variation and change. 

Adolescent speakers were identified as a major social group for linguistic variation and change 

(Kerswill, 1996; Tagliamonte, 2016; Smith & Holmes-Elliott, 2022). Nagy (2017, p. 35) 

summarizes the widely-in-press sociolinguistic generalizations regarding age (the references to 

the authors below are given by Nagy, 2017, p. 35): 

• Younger speakers show newer phases of variety than older speakers (Bailey et al., 1991) 

• Peak of innovations is reached at the adolescent age (D’Arcy, 2013, p. 490) 

• Vernacular reorganization of language by younger speakers (Labov, 2001) 



Chapter 2. Heritage Speakers 

16 

• Language stabilizes during adulthood (Walker & Meyerhoff, 2013, p. 178; Sankoff, 2013, 

p. 261, but counter-examples by D’Arcy, 2013, p. 489 and Sankoff, 2013, p. 270) 

As for the HSs, it has been shown that HSs of different age groups differ from age-matched 

monolingual speakers. For example, Alexiadou et al. (2021) examined gender agreement 

mismatches in the narrations by HSs of Greek in the US and monolingual speakers of Greek. 

Speakers of two age groups were included: adolescents (N = 32 in each group; mean age = 16.2 

for HSs and 15.3 for monolinguals) and adults (N = 32 in each group; mean age = 30.2 for HSs 

and 27.4 for monolinguals). All participants were tested in language situations that were distinct 

regarding formality (formal vs. informal) and mode (spoken vs. written). The results of the study 

revealed that HSs of both age groups, unlike monolingual speakers, showed mismatches in gender 

agreement. Formality and mode did not have an effect on the production of gender mismatches; 

however, there were significant differences observed between HSs of different age groups. 

Specifically, adult HSs showed gender agreement mismatches less frequently than adolescent HSs 

(Alexiadou et al., 2021).  

The differences between adolescent and adult HSs are explained with reference to the literacy 

practices in Greek (i.e., listening to music/radio/audiobooks, watching TV/movies/videos, texting, 

writing emails or blogs). The regression analysis revealed the following significant difference 

between adult and adolescent HSs: adult HSs were found to have higher literacy practices in Greek 

than adolescent HSs (Alexiadou et al., 2021). However, no correlation was found between the 

frequency of gender agreement mismatches and literacy practices. The input in HL as well as 

visits to Greece did not correlate with the frequency of gender agreement mismatches either. A 

weak positive correlation between the frequency of total productions and literacy practices was 

identified, i.e., adolescent HSs with significantly lower literacy practices in Greek than adult HSs 

were also found to produce significantly fewer overall instances of agreement (Alexiadou et al., 

2021). 

Alexiadou et al. (2021) also provide some preliminary results of the mismatches in gender 

agreement produced by HSs of Greek residing in Germany (adolescents: N = 21, mean age = 16.5; 

adults: N = 27, mean age = 28.4) although the full analysis of the German group has not yet been 

carried out since the data collection was still ongoing. The preliminary results suggest that HSs in 

Germany of both age groups show fewer mismatches in gender agreement than HSs in the US. 

Further, there seems to be no difference between adolescent and adult HSs in Germany (Alexiadou 

et al., 2021). Thus, the differences between HSs of different age groups may not be similarly 

pronounced in HSs with different MLs. 



 

17 

Sometimes younger HSs were found to be more similar to monolingual speakers of their HL 

than adult HSs. For instance, Polinsky (2011) investigated the comprehension of subject and 

object relative clauses by HSs of Russian in the US and monolingual speakers of Russian. The 

speakers belonged to two age groups: children (mean age: 6, 6 years old for monolinguals and 6.2 

for HSs) and adults (mean age: 32 years old for monolinguals and 22 for HSs) (Polinsky, 2011, p. 

316). The results of the study revealed that children of both speaker groups behaved similarly to 

each other while adult HSs differed from adult monolinguals. Specifically, adult HSs showed 

significantly weaker results than adult monolinguals in interpretation of object relative clauses but 

not subject relative clauses. The findings of the study were interpreted as language attrition.  

It has also been found that depending on an investigated phenomenon, HSs of different age 

groups may differ from monolinguals or be monolingual-like. A study by Van Osch et al. (2019) 

examined the knowledge of word order in intransitive sentences by HSs of Spanish living in the 

Netherlands. HSs of three age groups were investigated: children (circa 9 years old), early 

adolescents (approximately 13 years old), and adults (from 19 to 36 years old, mean age = 25). 

The data of HSs were compared to age-matched monolingual speakers of Spanish. The speakers 

were tested regarding their sensitivity to the verb type, focus, and definiteness (these constraints 

are relevant for the word order choice with intransitive verbs in Spanish) and their preference for 

the preverbal or postverbal position of subjects.  

The results of the study showed that monolingual speakers of all three age groups were 

sensitive to all three factors that are relevant for the word order, namely verb type, focus, and 

definiteness. As for the HSs, HSs of all age groups were sensitive to the verb type. The results for 

focus and definiteness were different for HSs of different age groups. Younger HSs (both children 

and young adolescents) were not sensitive to the factor “focus” while adult HSs were. This result 

shows that the effect of focus on word order eventually forms part of the HSs’ grammar (Van 

Osch et al., 2019). As for the effect of definiteness, younger HSs (both children and young 

adolescents) were sensitive to this factor while adult HSs were not. This result shows that the 

knowledge of some phenomena can become weaker later in life (Van Osch et al., 2019). Results 

regarding the overall word order preference in intransitive sentences reveal the following. 

Monolingual children showed a preference for SV order while young adolescent and adult 

speakers did not show any preference for a particular word order. As for the HSs, younger HSs 

(children and adolescents) did not show any preference for a particular word order while adult 

HSs showed preference for the VS order (Van Osch et al., 2019). The results of the study are 

explained with reference to the reduced input and the use of the HL together with the increased 

input and use of the ML. 



Chapter 2. Heritage Speakers 

18 

Polinsky (2016) summarizes the findings of different studies related to HSs of different age 

groups (children and adults) regarding various outcomes. Some studies showed that child HSs, 

adult HSs, and monolingual speakers were similar to each other regarding some phenomena (e.g., 

Polinsky, 2016, p. 550). Some research reported that child HSs pattern with monolingual children, 

but adult HSs do not pattern with adult monolinguals (e.g., Polinsky, 2011). Another possible 

scenario of age-related differences is that child HSs differ from monolinguals while adult HSs are 

similar to monolingual speakers (Polinsky, 2016, pp. 555-557). 

Adult HSs were sometimes found to show some similarities with young L1 children (e.g., 

Sekerina & Sauermann, 2015). However, it was also found that adult HSs differ from the baseline 

while young L1 children do not (e.g., Polinsky 2011). Furthermore, in some language domains, 

HSs were found to be similar to the baseline while L1 children were not (e.g., Arslan et al., 2015). 

Adult HSs were also sometimes found to show some similarities with L2 learners, another 

group of bilinguals. Such similarities were found in areas where the language transfer is quite 

strong: word order, lexical expressions (calques or translations from the dominant language), the 

use of overt elements, or the absence of infrequent forms (Polinsky & Scontras, 2019, p. 3). 

As it can be seen adult HSs may be closer to monolingual speakers of their HL, contributing 

data from a linguistically more stable age (e.g., Alexiadou et al., 2021 on HSs of Greek in the US). 

At the same time, adults might be more dominant in their ML and have less contact with their HL 

(since they typically do not live with their families anymore), and consequently they may be more 

different from monolingual speakers of their HL than younger HSs (e.g., Polinsky, 2011). The 

same holds for adolescent HSs: they may be similar to monolingual speakers of their HL (e.g., 

Van Osch et al., 2019 on the effect of definiteness on word order), or they may differ from 

adolescent monolingual speakers (e.g., Van Osch et al., 2019 on the effect of focus on word order). 

Additionally, the exact linguistic phenomenon may influence the performance of HSs at different 

ages as shown by Van Osch et al. (2019). 

2.3.2 Gender 

Age is not the only important parameter for linguistic variation and change. Along with age, 

gender was reported to be among the main social categories crucial for language change. 

Specifically, female adolescent speakers were found to be a driving force of linguistic change 

(Tagliamonte, 2016, p. 43). Nagy (2017, p. 35) summarizes the widely-in-press sociolinguistic 

generalizations regarding gender: 

• Women use stigmatized variants less frequently than men (Bayley, 2013, p. 14) 

• Women are the innovators in linguistic change (Labov, 1990, p. 205) 



 

19 

• Women are similar to men of the immediately higher social class (Bayley, 2013, p. 14) 

Although some studies on HLs did not find an effect of gender in bilingual populations (Nagy, 

2017; Nagy & Brook, 2020) many studies on HLs report that gender may have an effect on the 

way HLs are used. For instance, a study by Portes and Hao (1998) investigated the proficiency of 

HSs of different HLs in the US. It was found that female students whose parents came from Latin 

America were more proficient in their HL than male students presumably because females spent 

more time at home and were exposed to their HL more often than male speakers (Portes & Hao, 

1998, p. 279). 

Gender differences have been frequently reported in studies on phonetics, phonology, and 

intonation in monolingual as well as bilingual populations. As for the HSs, Sheikhbahaie (2020) 

investigated the acoustic features of vowels – first and second formants – produced by HSs of 

Farsi residing in the US. Generally, HSs were similar to first-generation immigrants (who had 

lived in the US for about 25 years), and first-generation immigrants were similar to monolingual 

speakers of Farsi. There were gender differences regarding one vowel, namely /æ/. Female HSs 

and first-generation immigrants were similar to female monolingual speakers of Farsi while male 

HSs and male first-generation immigrants produced /æ/ that was relatively higher than the one of 

male monolingual speakers of Farsi. 

Ordin and Mennen (2017) investigated the differences in fundamental frequency (F0) range in 

the read speech of simultaneous Welsh-English bilinguals. The findings reveal an interesting 

effect of gender on the frequency range. Specifically, only female speakers, but not male speakers, 

showed differences in their F0 in Welsh and English by producing wider spans and higher F0 

maxima in their Welsh compared to their English. Such results are explained with reference to 

cultural and personal identity. Specifically, men are often raised to be more independent and 

competitive while women are often raised to be cooperative and follow the rules of society. Thus, 

when women produced distinctive F0 patterns in both Welsh and English they accommodated to 

each of the languages as opposed to men who behaved more independently by not changing their 

F0 values in different languages (Ordin & Mennen, 2017). 

Van Rijswijk et al. (2017) investigated different focus structures in the ML Dutch of HSs of 

Turkish residing in the Netherlands. It was found that HSs and monolinguals differed from each 

other the most with respect to the pitch range in broad focus. Specifically, Dutch monolinguals 

demonstrated declination whereas HSs did not; they had the same pitch level during the whole 

sentence which was consistent with the intonational system of Turkish (Van Rijswijk et al., 2017, 

p. 65). Furthermore, HSs of different genders were found to use different strategies to maintain 

pitch at the same level: female HSs produced a higher peak on the object compared to female 



Chapter 2. Heritage Speakers 

20 

monolingual speakers while male HSs produced a lower peak on the subjects than male 

monolingual speakers. Gender was found to be important for some other F0 differences between 

speaker groups. Female bilinguals differed from female monolinguals by producing higher peaks 

on the subject and object in every tested condition while male bilinguals differed from male 

monolinguals by producing lower peaks on subjects and objects than monolingual male speakers.  

Apart from the difference between gender and speaker groups, there was an effect of gender 

on its own. Male speakers of both groups produced a larger fall on the object compared to female 

speakers of both groups.  

Additionally, male bilingual speakers were found to differ from other speaker groups by 

producing a similar fall on the object in broad and contrastive focus as opposed to other groups 

that produced a steeper fall on the contrastive focus compared to other focus conditions (Van 

Rijswijk et al., 2017, p. 64). 

The results of the gender differences are explained with reference to two factors. The first factor 

has to do with culture and society. For instance, both male speaker groups produced steeper falls 

on the object than female speakers and such falls were reported to be associated with self-

confidence and masculinity (Van Rijswijk et al., 2017, p. 65 citing Gussenhoven, 2005). The 

second factor that may explain the gender differences is related to the prestige and attitude towards 

the language. Male bilingual speakers did not mark the contrastive focus with a larger fall while 

all other speaker groups did. Male bilingual speakers possibly showed an influence from Turkish 

since in Turkish contrastive focus is not expressed by a large fall while female bilingual speakers 

did not show an influence from Turkish adopting Dutch intonation since Dutch is the prestigious 

language in the Netherlands. Such gender differences go hand in hand with previous findings 

about women often taking a leading role in language change and using the prestigious language 

variants (Van Rijswijk et al., 2017, p. 65 citing Labov, 2001; Simonet, 2011). 

Some studies found an effect of gender in one of the populations of HSs, but not in another 

population. For instance, a study by Altenberg and Ferrand (2006) investigated the mean F0 and 

the F0 range of two bilingual populations: HSs of Russian and HSs of Cantonese residing in the 

US. Participants were tested in their heritage and majority language in connected speech (i.e., 

participants were asked to describe their summer vacation). All the participants including 

monolingual speakers of American English were female speakers. The results of the study 

revealed that HSs of Russian produced significantly higher mean F0 in their HL than in their ML 

while HSs of Cantonese did not demonstrate a significant difference between their mean F0 in the 

HL and ML. Besides, both speaker groups behaved similarly to monolingual speakers of English 

regarding their mean F0 (Altenberg & Ferrand, 2006, p. 92). The results of the mean F0 are 



 

21 

explained with reference to the differences in language proficiency and linguistic and cultural 

characteristics. Most of the HSs of Cantonese were born in the US while the majority of HSs of 

Russian came to the US during their childhood between 4 and 13 years of age. Thus, HSs of 

Cantonese were supposedly less proficient in their HL than HSs of Russian. However, this 

assumption turned out not to be entirely plausible since one HS of Cantonese who was born in 

Burma did not show the higher mean F0 in Cantonese, but in English. Thus, Altenberg and Ferrand 

(2006) conclude that it is linguistic and cultural differences between two HS groups that contribute 

to the differences in the mean F0 rather than proficiency in HL. 

Gender and age were also found to interact. For example, monolingual speakers of Russian of 

different age groups and genders were found to show different linguistic behavior. Specifically, 

they were found to show differences in the frequency of null subjects depending on age and 

gender: younger speakers and male speakers produced null subjects less frequently than older 

speakers and female speakers (Nagy, 2015, p. 320).  

Simonet (2008, 2011) investigated utterance-final pitch accents (PAs) in the narrations of 

Catalan-Spanish bilinguals in Majorca who were either more dominant in Spanish or in Catalan. 

Different age groups (younger and older speakers) and genders were included as factors that could 

influence the production of utterance-final PAs. It was found that Spanish-dominant female 

speakers produced more accents in their Spanish that are typical of Catalan than Spanish-dominant 

male speakers. As for the productions in Catalan by Spanish-dominant speakers, it was found that 

female speakers produced PAs in their Catalan that were similar to the PAs of Standard Catalan 

while male speakers transferred PAs from their dominant Spanish into their Catalan. As for the 

effect of age, it was found that younger Spanish-dominant speakers produced more accents in their 

Spanish typical for Catalan while older speakers generally produced accents typical for Spanish. 

Overall, Simonet (2008) showed that the utterance-final PAs in Spanish were influenced by the 

age and gender of speakers. The speakers can be placed on the following scale, from more 

Spanish-like accents to more Catalan-like accents: “older Spanish-dominant males > younger 

Spanish-domin