Thesis - Semantic Scholar...Thesis Reference The acquisition of Jamaican Creole: The emergence and transformation of early syntactic systems DE LISSER, Tamirand Nnena Abstract This

Thesis

Reference

The acquisition of Jamaican Creole: The emergence and

transformation of early syntactic systems

DE LISSER, Tamirand Nnena

Abstract

This study explores the early acquisition of Jamaican Creole (JC) syntax. There is a

significant gap in linguistic research investigating the acquisition of creole languages which

this research aims to repair. Six children, age ranging from 1;6 – 1;11, were recorded over an

18 month period. 60 minutes recordings were conducted every 10-15 days, thereby

establishing the first longitudinal corpus of acquisition data in a creole language. The corpus

was subjected to detailed analysis describing both target-consistent and target-inconsistent

productions. The phenomena studied included the emergence of tense, aspectual and modal

markers, null subjects, focalization, topicalization, interrogation and ‘typical creole features'

such as verb serialization, double-negation, etc. The empirical findings provide evidence of

early syntactic development and contribute to the on-going debate on Language Universals.

This study will have a long-lasting contribution to the linguistic community as it provides an

accessible corpus of natural production of JC early syntactic systems.

DE LISSER, Tamirand Nnena. The acquisition of Jamaican Creole: The emergence and

transformation of early syntactic systems. Thèse de doctorat : Univ. Genève, 2015, no. L.

835

URN : urn:nbn:ch:unige-742285

DOI : 10.13097/archive-ouverte/unige:74228

Available at:

http://archive-ouverte.unige.ch/unige:74228

Disclaimer: layout of this document may differ from the published version.

1 / 1

http://archive-ouverte.unige.ch/unige:74228

The Acquisition of Jamaican Creole:

The Emergence and Transformation of Early Syntactic Systems

Doctoral thesis presented by:

Tamirand Nnena DE LISSER

2015

Thesis Jury: Professor Ur Shlonsky: University of Geneva (President)

Professor Luigi Rizzi: University of Geneva and Sienna (Director)

Doctor Stephanie Durrleman: Centre National de la Recherche Scientifique

Professor Hubert Devonish: University of the West Indies, Mona

Professor Dany Adone: University of Cologne

Professor Enoch Aboh: University of Amsterdam

i

For my son Kyahri

ii

TABLE OF CONTENTS

ABSTRACT……………………………………………………………………………..

ABBREVIATIONS……………………………………………………………………..

ACKNOWLEDGEMENTS……………………………………………………………

CHAPTER 1

Introduction……………………………………………………………………………..

1.1 Motivation………………………….…………………………………….....……

1.2 Goals of the study…………………..……………………………………….……

1.3 Organization of dissertation………….……………………………………..……

viii

ix

x

1

1

3

4

CHAPTER 2

Theoretical Background and Framework………………………………………….…

2.1 Introduction……………...…………………………………….…………………

2.2 General background and concepts………………….………….…………………

2.2.1 Basic concepts and ideas…………..…..…..…...…………………………

2.3 Theoretical framework…………………………………………………………...

2.3.1 Generative grammar..................................……..…………………………

2.3.2 Principles and parameters framework.....……………...……..…………...

2.3.3 Extra-grammatical explanations………...………………………………

2.3.4 Language bioprogram hypothesis………………………...……………..

2.4 Conclusions………………………..………………..……………………………

7

7

7

9

11

11

13

14

15

18

iii

CHAPTER 3

The Language Situation and Grammar……………………………………………….

3.1 Introduction………………………………………………………………………

3.2 Language situation…………………………………………………………..……

3.3 Jamaican Creole syntax……………………………………………..……………

19

19

19

23

CHAPTER 4

Methodology…………….………………………………………………………………

4.1 Introduction………………………………………………………………………

4.2 Participants…………………………………………………….…………………

4.2.1 Participants individual profiles……………………...……………………

4.3 Recordings………………………………………..………………………………

4.4 Transcriptions, coding and analysis…………………………......….…….…...…

4.4.1 Coding………………………….…………………………………………

4.4.2 Analysis…………………………………………………………………

29

29

29

33

39

41

42

44

CHAPTER 5

General Developmental Patterns in JC……………………………………………….

5.1 Introduction………………………………………………………………………

5.2 Developmental stages …………………………………………………………

5.3 Grammatical categories…………………………………………..………………

5.3.1 Utterance types……………………………………………………………

5.3.2 Word classes………………………………………………………………

5.3.3 Lexical categories…………………………………………………………

5.3.4 Functional categories………………………………………………..……

47

47

47

51

52

55

57

61

iv

5.4 Initial combinations………………………………………………………………

5.5 Multiword utterances……………………………………………………..………

5.6 Discussion & conclusion…………………………………………………………

5.6.1 Summary……………………………………………………….…………

63

68

73

74

CHAPTER 6

The acquisition of tense, modal and aspect…………………………………………...

6.1 Introduction………………………………………………………………………

6.2 An overview of tense, modal & aspect in JC…….………………………………

6.3 Modality………………………………………………………………………….

6.3.1 Acquisition of modality…………………………………………………...

6.3.2 Acquisition of modality in JC ………..…………………………………

6.3.3 Order of acquisition of modal markers…………………...………………

6.3.4 Omission of modals.........…………………………………………………

6.3.5 Intermediary conclusion………………………..…………………………

6.4 Tense……………………………………………………………...………………

6.4.1 Acquisition of tense……………………………………………………….

6.4.2 Acquisition of tense in JC…………………...……………………………

6.5 Aspect…………………………………………………………………………….

6.5.1 Acquisition of aspect……………………………………………………...

6.5.2 Acquisition of aspect in JC…………………………………….…………

6.5.3 Omission of aspect markers………………………………………………

6.5.4 Intermediate Conclusion……………………………….…………………

6.6 Root infinitives……………………………………………………………...……

6.6.1 Root infinitives in JC………………………………..……………………

76

76

78

80

81

81

90

92

94

94

94

95

99

100

100

109

115

116

117

v

6.7 Cumulative development of the TMA zone…………...……………….………..

6.7.1 Co-occurence of TMA markers………………..…………………………

6.7.2 Correlation of children’s utterance with input data……………………….

6.8 Discussion & conclusion…………………………………………………………

6.8.1 Summary of main findings……………………..…………………………

6.8.2 Discussion…………………………………………………...……………

6.8.3 Conclusions…………………………………………………….…………

6.8.4 Future Research……………………………………...……………………

125

128

134

136

136

137

142

142

CHAPTER 7

Null subject phenomenon………………………………………………………………

7.1 Introduction………………………………………………………………………

7.2 Object/subject asymmetry…………………………………..……………………

7.3 Early subject drop in JC……………………………….…………………………

7.4 Subject omission as a manifestation of the Privilege of the Root: overt wh-

questions………………………………………………………………………….

7.4.1 Overt wh-phrase and null subjects……………..…………………………

7.5 Null subjects with null question-elements……………………..…………………

7.5.1 Null wh……………………………………………………………………

7.5.2 Null wh-elements and null subjects in JC……………...…………………

7.5.3 Null subjects and yes/no questions in JC…………………………………

7.6 Revising the Privilege of the Root mechanism……………..……………………

7.7 Summary and conclusion……………………………………………...…………

143

143

144

153

164

167

175

175

177

184

187

193

vi

CHAPTER 8

The acquisition of wh-interrogation, focalization and topicalization……………….

8.1 Introduction………………………………………………………………………

8.2 Acquisition of Interrogation…………………………………………...…………

8.2.1 Acquisition of constituent questions in JC……….………………………

8.2.2 In-situ wh-phrase………………………………………………………….

8.2.3 Movement in early wh-questions in JC…………...………………………

8.2.4 Errors in production………………………………………………………

8.3 Acquisition of focus in JC…………………………………………..……………

8.3.1 Overt focalized fronted constituents and null subjects…...………………

8.3.2 Omission of overt focal marker a…………………………………………

8.3.3 Focalized fronted constituents with omitted focus marker and null

subjects……………………………………………………………………

8.4 Acquisition of topicalized constituents in JC …………………..………………

8.4.1 Target inconsistencies in early topicalized utterances……………………

8.5 Cartographic ordering of the left periphery in JC…………..……………………

8.6 Discussion and conclusion…………………………………………….…………

195

195

196

196

199

201

205

205

217

219

222

225

228

234

237

CHAPTER 9

The acquisition of typical creole features………………………………..……………

9.1 Introduction………………………………………………………………………

9.2 Negation………………………………………………………………………….

9.2.1 Multiple negation…………………………………………………………

9.2.2 Target inconsistency………………………………………………………

9.3 Serial verb construction………………………………………………..…………

239

239

240

249

252

253

vii

9.3.1 Accounting for serial verb constructions in creole acquisition………...…

9.4 Inclusive/plural marking…………………………………………………….……

9.5 Pronouns and reflexives…………………………………………….……………

9.6 Copula……………………………………………………………………………

9.7 Determiners………………………………………………………………………

9.8 Conclusions………………………………………………………………………

259

261

267

271

278

284

CHAPTER 10

Conclusions and implications………………………………………….………………

10.1 Introduction………………………………………………………………………

10.2 Summary and conclusions……………………………………………..…………

10.3 Implications for future work……………………………………..………………

286

286

286

288

APPENDICES…………………………………………………………………………..

Appendix 1: Consent form……………………………………………………………...

Appendix 2: Codes and grammatical relations of JC………….…..……………………

Appendix 3: Distribution of TMA in the input…………………………………………

REFERENCES………………………………………………………………………….

290

290

294

298

300

viii

ABSTRACT

This research explores the early acquisition of syntactic properties in Jamaican Creole (JC)

and their subsequent transformation into stable grammars. There is a significant gap in

linguistic research investigating the acquisition of Creole Languages. With the exception of

Adone’s (1994) and subsequent work, the acquisition of Creole Languages has not been

extensively investigated. Adone concluded that more cross-linguistic research is required to

provide more insights into the domain of creole language acquisition. The present research

seeks to fill this void by establishing a corpus of natural production of JC children, between

the ages of 18 – 40 months. This age range is not arbitrary as it corresponds to the period in

which syntax emerges in children, and where target-inconsistent forms and structures have

been documented in other languages (Guasti 2002; Radford 1990; Rizzi & Friedemann 2000).

Six children, from households where basilectal JC is the predominant language, were

recorded over an 18 month period. 60 minutes recordings were conducted every 10-15 days,

and the recorded material transcribed in accordance with the guidelines of the CHILDES

Database. The corpus was subjected to detailed analysis aimed as describing both target-

consistent and target-inconsistent productions. In an attempt to providing a broad overview of

the language, attention was paid to the emergence of tense, aspectual and modal markers, null

subjects, focalization, topicalization and interrogation, typical creole features such as serial

verb constructions, negative concord, etc. The empirical findings provide evidence of early

syntactic development and contribute to the on-going debate on Language Universals. This

study proves to be significant as it is the first longitudinal exploration of the acquisition of a

Creole Language. It will have a long-lasting contribution to the Linguistic community as it

provides an accessible corpus of natural production of JC early syntactic systems.

ix

ABBREVIATIONS

Ø Null

1SG 1st person singular pronoun

2SG 2nd

person singular pronoun

3SG 3rd

person singular pronoun

1PL 1st person plural pronoun

2PL 2nd

person plural pronoun

3PL 3rd

person plural pronoun

ABL Ability Modal

ASP Aspect

COM Communicator

COMP Completive Aspect

COP Copula

CMP Complementizer

DEF Definite Determiner

DEM Demonstrative

DET Determiner

EPIS Epistemic modal

EQU Equative Copula

FOC Focus particle

FUT Future

GEN Genitive

INCL Inclusive marker

INF Infinitive

INDEF Indefinite Determiner

LOC Locative

MOD Modal

NEC Necessity Modal

NEG Negator

NOM Nominative

OBL Obligation modal

PAST Past tense

POSS Possessive

PREP Preposition

PROG Progressive Aspect

PROS Prospective Aspect

Q: indef Indefinite Quantifier

REF Reflexive

REL Relativizer

RETRO Retrospective Aspect

x

ACKNOWLEDGEMENTS

This journey has brought me to the realization of how truly blessed I am, and I am happy that

there are many persons to be acknowledged for my countless blessings!

This research project was made possible due to funds received from the Swiss National

Science Foundation, to which I am grateful. I however owe my utmost gratitude to the Project

Applicants: Ur Shlonsly, Luigi Rizzi and Stephanie Durrleman. Firstly, I want to thank

Stephanie Durrleman, not only for envisioning this research, but more so for always being

there for me. Stephanie has not only been an advisor providing feedback and directives on my

academic life, but has been like family away from home. She has opened her doors to me and

I have benefitted tremendously from her warm hospitality and thoughtfulness. For these and

many more that I will not mention here in the interest of space and time, I am profoundly

indebted to her.

Next I would like to express gratitude to my Thesis Director, Luigi Rizzi, and the President of

my Jury, Ur Shlonsky, for their invaluable advice, directives, comments and feedback in

moulding me to produce this work. I could not have done it if it was not for their guidance,

encouragement, support and commitment. The opportunity afforded to work with them and to

benefit from their vast knowledge and experience is such an honour to me, of which I am

extremely grateful.

I would also like to thank Hubert Devonish, Dany Adone, and Enoch Aboh, who despite their

packed schedules have agreed to serve as members of my defence committee. I recently met

both Dany Adone and Enoch Aboh and have discussed with them various aspects of my

research. Their useful suggestions and general advice are greatly appreciated. It is indeed a

privilege to have them on my defence committee. My gratitude for Hubert Devonish is more

xi

time-honoured. Hubert Devonish introduced me to Linguistics, and has never faltered in

providing guidance throughout my maturity in this aspect of my life. ‘Prof’, as he is

affectionately called, would go the extra mile for his students, in ensuring our success. He is

always willing to brainstorm with us in finding the perfect solution to any problem that we

may face. I am at a loss for words that can justly describe the magnitude of thanks that I

would love to express to Hubert Devonish.

Other key persons to whom I express gratitude are the research participants and their families

who have welcomed me wholeheartedly into their homes and personal lives. Without their

commitment this research would not have been possible. I am extremely appreciative of them.

Immense credit is due to my Research Assistant, Nickesha Dawkins. The data collection and

transcriptions could not be realistically achieved if it was not for her dedication and

assistance. Furthermore, the general support and words of encouragement received from

Nickesha throughout this entire journey is dearly valued. Additionally, I extend my thanks to

Danielle Smith, Sheneil Ellis, Patrice Clarke and Tahirah Charles who all worked assiduously

in coding the data for analysis. The analysis could not be feasible without their diligence.

I would also like to say a special thanks to Per Baumann who had kindly accepted the

challenge of thoroughly proof-reading my entire dissertation and providing instrumental

feedback in improving the overall quality of my work. Merit is also extended to my office

mates Christopher Laenzlinger, Gabriela Soare, Goljihan Kashaeva and Tomislav Socanac for

their assistance and words of encouragements. Other members of the department that I

express gratitude to are Margherita Pallottino, Tabia Ishane, Genoveva Puscas, Jean-Philippe

Goldman, Annamaria Bentea, Frederique Berthelot, Giuliano Bocci, Paola Merlo, Sarah

Ouwayda and Jacques Moeschler. To all the other members of the department who have

provided assistance and or encouragement, I duly thank you. Of course this section would not

xii

be complete without a special thanks to the Department's Secretary, Eva Capitao. Eva is to be

acknowledged on so many levels; her willingness to assist with every request, her warm

embraces, her gifts, the chats which contributed to improving my level of French, and above

all her encouragements.

I would also like to take the time out to recognise the members of the Doctoral Schools which

I have attended and the audiences at SCL-SPCL-ACBLPE 2014, GALA 2013, FACS4 2014,

GRGC 2014, SPCL-ACBLPE 2013 and GDRI Meeting 2013 who have provided support and

feedback to my work. The Academic Society of Geneva and the SPCL Peck Fund should be

highlighted for their assistance in making my attendance at the SCL and SPCL meetings a

reality. Likewise I would like to express my profound gratitude to the Ernst et Lucie

Schmidheiny Fondation for the grant received.

Acknowledgements are also extended to a few friends who have provided support to me at

one point or another throughout my entire journey: Salwa Soliman, Florent Duponcel,

Tasheney Francis, Marcia Allen, Auline Smith, Sharon Henry, Carl Brown, Shane Lawson,

Rohan Whyte and the JAS Family. I would also like to thank my siblings: Lecia, Thirdis,

Jadine, Iley, Tibab, Ras Emosep, Kuwaame and Dennis. Their interminable love and support

is appreciated. Though we are oceans apart, at no time have I ever felt like I am in this alone. I

could not have done this without them. Finally I would like to thank the two most

instrumental persons in my life who have sculpted me to be who I am: my mother Blossom

and my son Kyahri. I love them and I will forever be grateful to them for believing in me!

I put forward gratitude to all the positive energies in the universe that have attracted other

positive energies, and in so doing, have engulfed my world. And for those who I have not

mentioned, please find comfort in the fact that you too have contributed to the pool of positive

energies that surrounds me and as such is duly recognised. I am truly grateful of my blessings.

Chapter 1 : Introduction

1

CHAPTER 1

INTRODUCTION

1.1 MOTIVATION

First language acquisition refers to the process by which a child develops his or her first

language. This is interesting not only from a theoretical perspective, but is a remarkable

experience for anyone, especially parents, observing the transformation of their children’s

linguistic achievements. Just yesterday my son was babbling, and before you could imagine

he is a ‘talking machine’. Where does this language come from? When exactly did it happen?

How do children acquire target-like competence in their linguistic production?

In the last 40 years, there has been increasing scientific research on these mysteries of

language acquisition. Recent research has shown that early syntactic processes obey

systematic patterns constrained by specific rules. Arguably, these rules are endowed by an

innate language mechanism which appears to be universal. These findings are however based

on the acquisition of European languages such as French, English, German, etc. It is just

recently that research in the domain of child language acquisition has been extended to non-

Indo-European languages such as Japanese, Korean, Chinese, etc. Still the acquisition of other

types of languages such as Creole languages remains a domain that has not been extensively

investigated. Apart from Adone (1994, 2012 and subsequent unpublished work), Adone &

Vainikka (1999), and Pratas & Hyams (2009), the acquisition of Creole presents a lacuna in

the field of language acquisition. To date, there has been absolutely no published work on the

acquisition of Jamaican Creole Syntax. Adone (1994:144) concludes that “more cross-


2

linguistic data is required to provide more insight into this domain together with the study of

Creole acquisition.”

The time is therefore ripe for more such research, as Bickerton (1999a:67) underlines when

considering the state of affairs: “The all-but-complete absence of studies of how creoles are

nowadays acquired constitutes a significant gap in our knowledge of acquisition, given the

extent to which our knowledge has expanded crosslinguistically in the past couple of decades

(Ochs 1982; Schieffelin 1979; Slobin 1985; etc.), even if creoles are no more than just another

ordinary language family. Given the way in which they are, in fact, formed, their significance

can surely not go much longer unexplored by acquisitionists.” The investigation of JC

acquisition is therefore a starting point in our quest to fill this void that exists due to the dearth

of studies on the acquisition of creole languages. Undertaking a study of the acquisition of JC

syntactic structures will address theoretical issues related to the field of language acquisition

at large.

Additionally, not only will this research contribute to scientific knowledge, but in the context

of Jamaica, it may be applicable to well needed developments in language planning,

particularly with regards to early childhood education. This study looks at the acquisition of

grammatical structures of JC and is hoped to contribute to general policy making for early

education, for example by helping teachers to have an idea of what the norm is for a child at

various ages and what is deviant. Moreover, being that English is the official language used in

education, the present study will not only reinforce the fact of the distinctiveness of JC and

English, but more importantly it will establish the plausible differences in the course of

acquisition of the two languages, thus providing a base for JC speakers’ inconsistent

productions of the language used in the formal school system.


3

Meade (2001) offers a descriptive account of the acquisition of Jamaican phonology, which

represents stages of segmental acquisition for children aged one to four and a half years. More

recently, Karla Washington, Sharynne McLeod, Hubert Devonish & Maureen Samms-

Vaughn (personal communication) have been doing Speech Language Pathology related

research covering the linguistic development norms of young children (aged three to six

years) in Jamaica. Along similar lines, the present research practically provides a much-

needed accessible database for further study of the Jamaican language situation and the

overall acquisition of the language.

1.2 GOALS OF THE STUDY

Based on longitudinal naturalistic data, this dissertation focusses on the emergence of the

syntactic systems of JC and investigates their transformation into the target grammar in the

course of the first years of life. The dissertation, with reference to relevant theoretical

concepts, provides an extensive descriptive analysis of both target-consistent and target-

inconsistent production in the development of the grammatical representations of the children

acquiring JC. The main goals therefore are:

a) To map the process of development, including time-lines and sequence, by which

children acquire specific grammatical structures.

b) To develop an accessible database of longitudinal naturalistic JC data.

c) To fill the gap in first language acquisition studies with respect to creole languages.

Creole languages, being of the analytic type, provide overt realization of various syntactic

elements and positions by using free morphemes. As such they are particularly suited for

mapping the incremental development of child grammar. As previous research on the


4

acquisition of such languages is sparse, this dissertation aims to provide insights to the many

questions that are still open:

1) Is there a developmental order in the acquisition of lexical and functional structures,

with the latter globally delayed with respect to the former (Radford 1990)? Or do

lexical and functional structures co-occur at all levels of acquisition?

2) Does structure emerge incrementally in line with the incremental structure building

approach to development (Radford 1990, and subsequent works) or are all structures

available when significant production starts in line with the full competence approach

(Poeppel & Wexler 1993)?

3) Does the grammar of children acquiring JC replicate the highly structured

cartographically-coherent pattern of the target language?

4) Is the acquisition of JC exceptional or is it acquired just like other well-studied

languages? Do learners of JC go through a root null subject phase? Does their

grammar reflect Root Infinities? Are structures acquired effortlessly and without

errors reflecting immediate target consistency?

In this thesis an attempt is made to evaluate competing accounts of early syntactic

development and add to the discussions of linguistic universals.

1.3 ORGANIZATION OF DISSERTATION

The dissertation is structured as follows. Chapter 2 provides succinct contextual details on the

state of scientific research in the field of first language acquisition and defines the main

concepts which are relevant to the present study. It then discusses the theoretical framework

on which the research is based.


5

Chapter 3 deals with background aspects of the Jamaican language situation which have

major impacts on the current study. This includes a description of the creole continuum and a

synopsis of language use in early education. Additionally, this chapter provides a concise

description of the syntactic features of the adult language which is the target structure for the

children studied in the present work.

Chapter 4 discusses the methodology. It gives details of the participants, their environments

and the criteria employed in their selection. Data collection procedures for gathering the

naturalistic speech data on which this dissertation relies are described. Additionally, the

procedures of transcribing, coding and analyzing the data are outlined.

Chapter 5 provides a step-by-step descriptive view of the general development patterns of

clauses thereby proposing an analysis of phrase structure in the initial phase of language

acquisition. It charts a system of grouping which enables a comparative analysis of early

grammatical development in terms of stages and maps the changes in acquisition as children

move towards the target system. It examines utterance types, lexical and functional

grammatical categories, and the basic principles of phrase and sentence formation as children

move from single-word to multi-word utterances.

Chapter 6 reports on the acquisition of inflection, focusing on the production and omission of

Tense, Mood and Aspect (TMA) markers. It examines what appears to be evidence of a Root

Infinitive Stage in early JC. It outlines the co-occurrence of various TMA markers thereby

establishing a cartographic representation of the Inflectional Phrase in early JC.


6

Chapter 7 provides a systematic analysis of early subject omission in JC. It discusses the

object/subject asymmetry and details the general pattern of the null subject phenomenon in

the production of the children in the corpus. The findings provide empirical support for the

“Privilege of the Root” and for the Truncation Hypothesis. The phenomenon of subject

omission in null wh-constituent questions and yes/no questions is best captured by a revision

of the approach to the “Privilege of the Root” and Truncation, along the lines of the spell-out

mechanism of Phase Theory.

Chapter 8 looks at the emergence of left-peripheral movement in children’s grammar,

focussing on movement operations involved in wh-interrogation, focalization and

topicalization. It examines the omission of overt focus markers and details the phenomenon of

null subjects in these environments. Additionally it provides a cartographic mapping of the

relative order attested in the development of the hierarchal structure of the Complementizer

Phrase.

Chapter 9 examines the acquisition of various constructions generally characterized as being

typical creole features and evaluates their development in terms of errorless acquisition. It

describes the early development of negation and negative concords, verb-serialization,

pluralization/inclusiveness, pronouns and reflexives, determiners, and the production and

omission of the copula.

Finally, Chapter 10 embeds the acquisition findings as presented in the previous chapters into

a brief discussion of the overall emergence and transformation of early syntactic systems of

JC, and concludes with remarks on future prospects.

Chapter 2: Theoretical Background and Framework

7

CHAPTER 2

THEORETICAL BACKGROUND AND FRAMEWORK

2.1 INTRODUCTION

There are some striking similarities in early syntactic developments which have been

observed cross-linguistically and for which varying theories of language acquisition have

been posited. The goal of this chapter is to provide a concise theoretical backdrop of the state

of scientific research in the field of first language acquisition (section 2.2) and to outline the

framework in which the present research is couched (section 2.3).

2.2 GENERAL BACKGROUND AND CONCEPTS

There are some basic questions which studies on first language acquisition generally aim to

answer: How do children systematically acquire language? Is there a strict developmental

order to language? What factors affect linguistic development? Various methods have been

employed in studies aimed to answer these questions; despite the differences in

methodologies, all studies converge on the fact that any human child can acquire the language

of their linguistic community effortlessly, without explicit teaching, on the basis of positive

evidence, under varying circumstances, and in a limited amount of time. In addition, despite

great differences in input and in conditions of acquisition, this process is achieved in

remarkably uniform ways cross-linguistically. For example, at about 6 - 8 months all children

start to babble; at about 10 - 12 months they speak their first words and between 20 and 24

months they begin to put words together (Guasti, 2002). Also the structures that children form

are similar across languages, whether they are target-consistent or not: for example research


8

shows that children in the initial stages of language acquisition omit sentential subjects,

regardless if these are features of the target language. Additionally, it is observed that by the

time children begin to utter word combinations equivalent to full sentences, they have already

figured out the rules governing word order and the position of verbs in the sentence. Thus the

production of a child acquiring English will manifest Verb-Object (VO) order while that of a

child learning Japanese will manifest Object-Verb (OV) order. French learners position the

inflected verb to the left of the negative particle pas while English learners position the

negative particle before the verb, as in the corresponding target language (Pierce 1992).

The vast majority of these results in acquisition studies, as reported above, originate from

research on the acquisition of major European languages. It is quite recent that research on

non-European languages (such as Japanese, Korean, Hebrew, and others) has become

significant. What is still lacking however, as alluded to in the Introduction, is research on the

acquisition of creole languages. It is important to include different types and families of

languages such as creoles in order to strengthen theory-guided research in acquisition.

Despite the languages being studied or the methodology and framework adopted, there are

some basic concepts which most syntactic theories of first language acquisition have adopted.

Some of these basic concepts underlying research in the domain of first language

development will be now discussed.


9

2.2.1 Basic Concepts and Ideas1

Language entails a psychological system realized in our mind/brain called a grammar. Our

linguistic knowledge allows us to produce an infinite set of utterances from a finite set of

lexical items, and also to understand sentences that we have never heard before. Additionally,

it gives us a tool to assess the acceptability of such utterances in accordance with the rules of

our language and the associated interpretation or meaning of the utterance. In short, grammar

is a system which ascribes certain structural representations to sentences and sanctions certain

interpretations while forbidding others. It does this by means of constraints that establish

whether certain constructions are possible or not possible in a language. Constraints are

linguistic principles that prohibit certain arrangements. Sentences that are considered well-

formed or acceptable in a particular language must conform to the linguistic constraints

governing the language. But how do children gain this grammatical knowledge?

Different hypotheses have been advanced, one of which holds that children learn language

through imitation (see Fraser et.al, 1963). The general idea here is that children imitate what

adults say by trying to repeat what they hear. However several facts showing that there is no

necessary similarity between the linguistic input that children receive and their subsequent

linguistic output argues against this hypothesis. For example children produce novel

utterances that they have never heard because the adult speakers in their environment do not

produce them. So an English learner would over regularize irregular past tense verbs, produce

target-inconsistent negation, utter ill-formed questions, etc. Moreover, children hear a finite

number of sentences, however they are able to produce and understand many sentences,

including those that they have never heard before and therefore cannot be imitating. Another

1 Ideas and discussions from this section are mainly from (Guasti, 2002).


10

idea is language learning through reinforcement. According to this view, children learn

language because they are positively reinforced when they produce correct utterances and

negatively reinforced when they make errors. This is in-line with one of the earliest scientific

explanations of language acquisition, known as behaviourism (Skinner 1948). The main idea

here is that children learn by echoing the behaviour of others, and when done correctly, is

positively reinforced. This hypothesis however, like the acquisition through imitation

hypothesis, cannot explain how humans acquire language and cannot characterize human

linguistic competence. First, it cannot explain how children acquire competence over an

indefinite number of sentences for which no reinforcement was provided. Second, in parent

child discourse, parents do not normally check for grammatical correctness, and if they should

do so, such corrections go unnoticed. Reinforcement therefore does not explain human

linguistic attainment.

The answer that Chomsky (1959 and much subsequent work) gave is that this grammatical

knowledge is based on an inborn predisposition. Arguments from the poverty of the

stimulus, that all speakers of a language know only a fairly abstract property and that this

property cannot be induced from the evidence available to children (positive evidence), point

to the requirement of an innate language mechanism. The hypothesis that the language

capacity is inborn and richly structured explains why language acquisition is possible, despite

all limitations and varying circumstances in which language learning takes place. It also

explains the parallels in the time course and content of language acquisition. Though

languages are similar, in accordance with this innate mechanism, there are very obvious

variations. These variations are however not unlimited, but are restricted by Universal

Grammar (UG). UG is the name given to the set of constraints with which all humans are

endowed at birth and that are responsible for the course of language acquisition.


11

Based on this background we will now examine the theoretical framework that guides the

present research.

2.3 THEORETICAL FRAMEWORK

This study is couched within the framework of Universal Grammar (UG) (Chomsky 1981a),

guided by the Principles and Parameters approach to language development. The theory of

UG proposes that there are common properties that all natural human languages share, which

are manifested innately without any formal training. Despite the varying circumstances and

the relatively limited amount of input data to which children are exposed, linguistic

milestones are achieved in a parallel fashion across languages. The resulting system of

knowledge is referred to as the grammar of the language the child is being exposed to.

According to Crain & Thorton (2006) it is UG which provides the roadmap for which this

course of grammar formation is derived. This roadmap consists of principles of grammar,

unvaried and valid for any possible human language, and a restricted set of options called

parameters, which together define the range and limits of possible cross-linguistic variation.

From this angle, language acquisition therefore consists of fixing the values of the various

parameters (Chomsky 1981, Hyams 1986, Wexler 1998).

2.3.1 Generative Grammar

The study of generative grammar is basically concerned with determining and characterizing

the linguistic capacities of individuals, both the initial state that is a common human

endowment and the subsequent grammar attained by the language faculty. In the generative

framework, work in the field of language acquisition tends to fall in two main categories:


12

continuity and maturation. According to the continuity model, grammatical principles are

available at all stages from birth and do not change (Pinker 1984, Borer & Wexler 1987).

Early grammar has all the same properties of the adult language, and the gap between early

and adult systems is accounted for by the assumption that children must learn language-

specific properties, and that development results in growth in other domains such as the

lexicon, pragmatic competence and processing abilities. Radical versions of the continuity

hypothesis, generally referred to as the Full Competence hypothesis allows for the least

degree of freedom for the child’s grammar to diverge from the adult’s, and as such relies on

external factors to account for changes in development (as discussed by Rasetti, 2003).

In the maturation model (postulated by Radford 1990; Vainikka 1993/4; Borer & Wexler

1987; Wexler 1998; among others) the basic assumption is that there is a genetically

determined maturation of grammatical categories and principles initially absent from child

grammars. The gap between early and adult systems is accounted for by the immature state of

UG. Although UG is available at birth, not all aspects are readily available and developmental

time may be extensive (Borer & Wexler, 1987). The basic idea of maturation-based

approaches is that children systems contain only the basic structure of UG, which is modified

over time by eventual changes and additions of specific properties compatible with UG.

Some proponents of the full-competence hypothesis assume maturational processes that may

impact language development. For instance, Rizzi (1992, 2000) assumes that children can

initially produce truncated sentence structures without some functional projections because

the constraint that requires all sentences to be full CPs only matures later2. These truncated

2 But see Rizzi (2005) for an analysis of truncation as a parametric property.


13

structures are not arbitrary, but grammatically based, and child grammars remain highly

constrained by UG.

2.3.2 Principles and Parameters Framework

Under the Generative Framework, Chomsky and Lasnik (1993) posited that in the normal

course of development, the language faculty passes through a series of states in early

childhood, from an initial genetically-determined state (UG) to a relatively steady state that

undergoes little subsequent change (target grammar). It is assumed that UG specifies an

infinite range of symbolic objects (referred to as structural descriptions) for each of the

permissible phonetic forms for human languages. This initial specification operates in much

the same way in every child and is rich enough to account for the attainment of a specific

language on the basis of the linguistic evidence available to the child. Arguably, it is not

readily operative in its full-fledged form but is learnable through experience, and is in fact an

empirical discovery for the child. Linguists working in the Principles and Parameters

framework therefore aim to determine the nature and degree of the grammatical knowledge

available to the child at specific stages of development. While some proponents argue that

linguistic abilities are observed to develop overtime, and children’s grammars appear to

transform with development, there is still a debate of whether early grammatical knowledge is

entirely available through all stages of development or whether some parts of it increase

overtime.

It is hypothesized that in the course of language development, children will need to “factor

out” overarching principles of UG that govern general rule application in order to access the

actual rules of the target grammar in the simplest form, with the relevant principles ensuring


14

that they will operate in such a way consistent with the input data available. Such ideas lead to

the principles and parameters approach. The principles and parameters hypothesis is that all

principles are assigned to UG and that cross-linguistic variation is restricted to certain options

as to how the principles apply. For instance, a directionality parameter was suggested to

capture the fact that phrases can either be left-headed (e.g. English VPs with VO-order) or

right-headed (e.g. Japanese VPs with OV-order). Likewise a pro-drop parameter allows for

the omission of subjects depending on what values apply to the language being acquired: e.g.

Italian is +pro-drop, where pronominal subjects can be dropped, while English is –pro-drop,

which does not allow pronominal subject drop. Under this approach, the theory of language

acquisition is then concerned with the acquisition of lexical items, fixing the values of the

parameters in line with the target language, and perhaps maturation of principles (Hyams

1986, Roper & Williams 1987, Borer and Wexler 1987, Crain 1991). It is argued that

grammatical “errors” committed by children in a non-systematic manner involve performance

errors while extremely systematic target inconsistencies are attributed to a mis-setting of some

parametric values. The distinction between the two however remains a moot and theoretically

relevant issue (Rizzi, 2006).

2.3.3 Extra-grammatical Explanations

Language Acquisition may be also based on and explained by extra-grammatical factors and

general cognitive development. One such approach to the study of acquisition has been

labelled the ‘functionalist approach’. The general belief here is that the grammatical forms of

any language are in some way determined by their communicative function and/or by features

of general human cognition (Ritchie & Bhatia, 1999). Semantic and pragmatic constructs are

integral to this approach to language development. For example, Greenfield & Smith (1976)


15

and Allen (2000) suggested that children tend to drop elements perceived as contextually

salient, i.e. the material most easily recoverable from context. In addition, other proponents of

extra-grammatical explanations argue in support of the ‘Classical Processing Approach’ to

child language acquisition. This system of development is characterized by the development

of performance or processing abilities (Bloom, Miller & Hood 1975; Bloom 1990; Valian

(1991); among others). The general idea is that the child system is not able to deal with the

grammatical complexity of the adult system, and as such deviations from the target (e.g. in the

form of omissions) are evident. Bloom et al. (1975) for example, relate utterance length to

factors ranging from word familiarity to discourse features. Additionally there is the Usage-

Based Theory of child language acquisition as advocated by Tomasello (2000, 2003).

According to Tomasello (2000:1) “most of children’s early linguistic competence is item

based, and therefore their language development proceeds in a piecemeal fashion with

virtually no evidence of any system-wide syntactic categories, schemas, or parameters.”

According to Rasetti (2003) “the central problem with performance-based explanations is

their vague and unrestrictive characterization of pragmatic and performance delays, which is a

consequence of their relatively underdeveloped status of research in these areas”. In a

nutshell, extra-grammatical theories cannot account for the developments of generative

grammar and the principles and parameters approach as posited in the current research.

2.3.4 Language Bioprogram Hypothesis

According to Bickerton (1981, 1984, 1999, 1999a), the idea of UG is supported by creole

languages because of the shared features among these languages. He argues that the structural

similarities among creole languages cannot be solely attributed to their superstrate (i.e.


16

European languages spoken by the dominant group during language contact) and substrate

(i.e. West African languages which were the native tongues of the slaves) languages, but

instead is a result of children’s innate language device. He refers to this innate device as the

language bioprogram, on which his theory, the language bioprogram hypothesis (LBH) is

based.

Much advancement has been made on the description of various developmental phenomena

after the LBH was first put forth. Bickerton (1999, 1999a) subsequently developed the

bioprogram theory and its current state will be set forth in this section. According to

Bickerton (1999, 1999a) the bioprogram consists of a set of default settings that apply when

no evidence of any other pre-existing system is incorporated into the grammar. It consists of a

finite list of options (features of languages that can be grammatically marked by morphemes

as opposed to lexical items) such as number, gender, person, tense, modality, aspect,

specificity, and so forth. It is assumed that there is a universal list of grammaticizable

distinctions that languages must choose from. In the event where no, or an inadequate number

of distinctions are present in the input (for example in creolization contexts), the child will

revert to a set of default distinctions. These default distinctions are the features that are

repeatedly found among creole languages (Bickerton, 1999).

If such a bioprogram for language exists, it is expected to play some vital role in the normal

acquisition of language by children acquiring non-creole languages. As pointed out by

Bickerton (1981) children acquiring English and other non-creole languages consistently use

certain structures in a target-inconsistent manner. These structures however are found to be

consistent with the rules hypothesized for the bioprogram and also the surface structures

found in creole languages. For example children acquiring English use negative subjects with


17

negative verbs (nobody don’t… ); zero copula (I playing); adjectives and intransitive verbs as

transitives and causatives (full my bucket, I’m gonna fall this on her); and other features

typical of creole languages. Thus if the child should have been learning a creole language, the

usage of such structures would be grammatical, reflecting immediate target consistency.

Bickerton (1999, 1999a) further states that children acquiring non-creole language may

quickly acquire non-Bioprogram features and not necessarily enforce the default-list

distinctions, or only employ them briefly where other universal grammaticizable distinctions

are readily available in the target language. However if the target-language possesses the

default creole-like features, these may well be acquired more rapidly and with significantly

fewer errors than features not found in creoles. He points out that if the choice of parameter

settings was open and made entirely on the basis of positive evidence; then children learning

creole languages should produce target-inconsistent structures with roughly the same

frequency as children learning non-creole languages. If however children acquire existing

creoles largely without errors, then the possibility that there is only a single non-

parameterized syntactic module, i.e. the default distinctions of the Bioprogram (and that all

parameter setting accordingly takes place in the lexicon) would be increased. This single

development is seen as providing all that is necessary for both the acquisition of pre-existing

grammars, and if necessary, for the creation of a novel grammar. This is contrary to

maturational models where different modules of syntax are envisioned as coming on line at

different times. Adone (1994) has found the acquisition of Mauritian Creole to be largely

errorless in the grammatical areas covered and thus strongly supports the LBH.

This proposal has clear similarities with Chomsky’s UG; however there are some differences

of details. Chomsky proposed a devise that potentially generates a wide range of possible


18

grammars, from which the child choose on the basis of primary input. Conversely, the LBH is

seen as a single, invariant innate grammar that might be subject to modification through

relevant experience. It is not derived from input but is derived from the language faculties of

the children concerned.

2.4 CONCLUSIONS

With the aim of offering a systematic and precise description of the development and

transformations that a child goes through in order to reach the stable target grammar, it is

necessary to adopt a comprehensible and structured framework which can produce testable

predictions and offer coherent guidelines for the analysis. For this reason, the present study is

embedded in, but not limited to, the parametric approach to language development and the

classical framework of UG. Alternate views to the study of language acquisition and the

general construction of grammatical knowledge will be considered whenever relevant. In

addition, this study aims to test some of the predictions of the LBH. Being that longitudinal

studies on the acquisition of creole languages are lacking, the present research will fill a void

in language acquisition studies.

Chapter 3: Background

19

CHAPTER 3

THE LANGUAGE SITUATION AND GRAMMAR

3.1 INTRODUCTION

This chapter deals with aspects of the Jamaican language situation which have major impacts

on the current study and presents a concise description of the syntax of the language. The

chapter is organized as follows: In section 3.2, I discuss the creole continuum involving three

varieties: the acrolect, the mesolect and the basilect, the latter being the focus of this

dissertation. The interaction between speakers of these varieties leads to significant linguistic

variation and overlap in their grammars. Additionally, I present a synopsis of the roles of the

languages in education, as some informants started to attend school in the latter stages of the

data collection phase. In section 3.3 I turn to the syntactic description of the basilect which

serves as a baseline of the target system that is being acquired.

3.2 LANGUAGE SITUATION

Jamaica, the largest English-speaking Island of the Caribbean, has a population of 2.7 million

inhabitants. Standard Jamaican English (English) is the official language, i.e. the language

used is schools, parliament and the media. Jamaican Creole (JC), an English-based creole,

descended from language contact between Africans and Europeans during and after slavery

(Roberts 2007), is the national language, and is spoken by the majority of the population. It is,

for the most part, the ambient language used in the home and is the first language of most

Jamaicans. Children acquire JC through parent, sibling and extended family interaction while


20

English appears to be primarily acquired from school interactions in the classroom (Carpenter,

2009).

The Jamaican language situation is generally described as a Creole Continuum (Decamp

1971) with speakers varying across the continuum from basilect to acrolect. At the basilectal

extreme, also referred to as the ‘deep creole’, speakers’ production manifests most substratum

influence (i.e. influence from West African languages as transmitted during slavery). This

variety is farthest from the ‘local standard’ and is generally associated with a rural setting.

Speakers at the other extreme (acrolectal end) are envisioned to speak the ‘local standard’,

which is the prestigious variety, containing the most ‘superstrate’ (British English) influences.

Situated in between the two extremes are the mesolectal varieties, which unequivocally share

attributes of both extremes in varying degrees. Speakers of opposite extremes may be

mutually unintelligible; however this is very rare as most people can adjust their variety

upward or downward on the continuum (Durrleman-Tame, 2008). Speakers of the basilectal

extreme are unintelligible to other Caribbean or foreign speakers of English (Lacoste, 2012).

The distinction between mesolect and basilect is not clear-cut as due to the relatively fluid

social structure, rural varieties are becoming more and more subjected to urban influence,

yielding more overlap between the two varieties (Winford, 1993). There are however,

speakers who command only one of the distinct varieties (monolinguals) and others who

command both varieties from either ends of the continuum (bilinguals). The Language

Competence Survey of Jamaica (2007) reports that 46.4% of its sample demonstrated

bilingualism; however only 17.1% were monolingual English speakers and 36.5% were

monolingual JC speakers.


21

The examples in (13) illustrate some possible variations in morphosyntax found across the

continuum (ranging from the basilectal extreme (a) to the acrolectal extreme (g)) for a single

statement.

1) a. Im wehn de nyam im fuud.

3SG PAST PROG eat 3SG food

b. Im behn a nyam im fuud.


c. Im did a nyam im fuud.


d. Im did a iit im fuud.


e. Im woz iitin im fuud.

3SG PAST eat~PROG 3SG food

f. Hii woz iitin im fuud.

3SG.NOM PAST eat~PROG 3SG.ACC food

g. Hii woz iiting his fuud.

3SG.NOM PAST eat~PROG 3SG.GEN food

‘He was eating his food.’

In this work, we concentrate primarily on the variety found at the basilectal extreme of the

continuum, which we have been referring to as JC. The choice for this selection is based on

the fact that it is the variety with the least influence from Standard English, and therefore

offering most syntactic novelty (in line with Durrleman-Tame 2008 and Bailey 1966).

3 Throughout this dissertation, examples which are not attributed a source are drawn from my native speaker’s competence. Note that I use the JLU modified Cassidy-LePage orthography (see section 4.4) for my own JC data, however where data is drawn from other sources, I use the spelling representations of those sources.


22

However, based on the continuum situation and the fact that English is the official language, it

is challenging to find speakers of only the basilect, having absolutely no access to the acrolect

or mesolectal varieties. Jamaicans, including those located at the basilectal end of the

continuum, would therefore exhibit some knowledge of English from early childhood, since

English is the variety used in education and the media (Lacoste, 2012). Due to the distinctive

social/communicative roles that the varieties fulfill, the Jamaican speech community has been

characterized as being diglossic (McCain, 1996). In the diglossic situations however, all

speakers possess some degree of competence in both the high and low varieties, and can

switch between varieties based on the situation of discourse.

During the course of the data collection phase some of the informants started to attend school.

It is therefore imperative to provide a sketch of the situation governing language use in early

childhood education. Children entering the school system in Jamaica would be predominantly

JC speakers, but due to the continuum situation there may also be a mixture with both JC and

English. As a result, The Language Education Policy (LEP) was developed in 2001 to

simultaneously promote oral use of JC in schools while facilitating the development of skills

in English (LEP, 2001). The Ministry of Education has adopted an approach in which teachers

“promote basic communication through the oral use of the home language in the early years

(e.g. Kindergarten to Grade 3) while facilitating the development of literacy in English”

(Bryan, 2001:23 in Lewis, 2010:13). This may include the teacher’s giving directions or

explaining a task in JC. Additionally, many teachers, especially at the kindergarten level, are

not fluent in English themselves. According to Bryan (2004), students and teachers may think

they are using English, when often they are not. So while the LEP upholds that English should

be used in the classroom, in reality, this is not necessarily the case. This may be due to the

lexical similarity found across the varieties along the continuum, and speakers not directing


23

attention to the grammatical distinctions between the two languages in operation. For example

JC is characterized by a cluster of grammatical properties which makes it quite distinct from

English, such as serial verb constructions, double negation, lack of subject-auxiliary

inversion, lack of case morphology or gender distinction on pronouns, etc. These features are

typically found in Atlantic Creoles (Patrick 2004) and will be examined in the subsequent

section.

3.3 JAMAICAN CREOLE SYNTAX

The grammar of JC has been comprehensively studied. Within the framework of generative

grammar, JC has been studied by Bailey (1966), Patrick (2004), and others who have worked

on specific grammatical properties of the language, such as Stewart (2006) on quantification.

More recently, Durrleman-Tame (2008) approached JC from the perspective of cartography

(Cinque 1999; Rizzi 1997). Based on (but not limited to) these detailed syntactic analyses, I

provide a general description of the relevant features of adult JC, which is used as the baseline

for target consistency throughout this dissertation.

JC is a Subject Verb Object (SVO) language as exemplified by the declarative clause in (2a).

Except in cases of predicate clefting and other kinds of fronting, other orders, such as SOV,

VOS, VSO, etc. as exemplified in (2b-d) are not felicitous as unmarked orders.

2) a. Kieti waan wan neda buk. (Bailey 1966)

Katie want one another book

“Katie wants another book.”


24

b. *4Kieti wan neda buk wahn.

Katie one another book want

c. *wahn wan neda buk Kieti.

want one another book Katie

d. *wahn Kieti wan neda buk

want Katie one another book

It is a non-null subject and a non-null object language and as such, it generally requires the

presence of overtly realized subjects and objects:

3) * Ø wahn wan neda buk.

Ø want one another book

“She wants another book.”

4) * Kieti wahn Ø.

Katie want Ø

“Katie wants it.”

Tense, modality and aspect (TMA) are expressed via free functional morphemes, typically

situated between the subject and the verb. These TMA markers and the verbal root remain

invariable for subject agreement.

5) Jan shuda ehn a ron. (Bailey 1966)

John MODAL PAST PROG run

“John should have been running.”

4 Throughout this dissertation * signals ungrammaticality.


25

6) Im wooda mos kyahn elp uno. (Durrleman-Tame 2008)

3SG MODAL MODAL MODAL help 2PL

“S/he would have necessarily been able to help you all.”

As seen in (5) and (6), these TMA markers can co-occur, however they must respect a

particular hierarchy, along the lines of the following (Durrleman-Tame 2008):

7) Mod epistemic > Tense > Mod root obligation > Mod root ability/permission >

Anterior > Asp continuative > Asp retrospective > Asp progressive > Asp

prospective > Asp completive > Asp frequentative

When non-stative verbs occur without TMA markers there is a default past reading, however

bare stative verbs are interpreted as non-past.

8) Mi kik Mieri.

1SG kick Mary

“I kicked Mary.”

9) Mi laik Mieri.

1SG like Mary

“I like Mary.”

JC makes use of Serial Verb Constructions (SVC) as in (10) below. These involve a series of

verbs, uninterrupted by a conjunction, which together express a single event with a single

subject.

10) Im tek mi aki sel a maakit.

3SG take 1SG ackee sell at market

“S/he sold my ackee at the market.”


26

There exists an equative and a locative copula in JC. The former is used to make parallel two

nominal elements as given in (11), and the latter precedes expression of location as in (12):

11) Jan a tiicha.

John COP teacher

“John is a teacher.”

12) Jan de skuul.

John COP school

“John is at school.”

The copula is however absent from certain constructions, for example, no copula is inserted

with adjectives. As such adjectival predicates behave similarly to stative verbs (Bailey 1966):

13) Jan fiesti.

John rude

“John is rude.”

Negative concord, the phenomenon where multiple negative elements together yield a reading

of negation, is attested in JC:

14) Im neva nyam notn.

3SG never eat nothing

“S/he never ate anything.”

Regarding interrogation, JC employs both simple bare wh-words such as uu ‘who’, wa ‘what’

and we ‘where’ and wh-compounds. Wh-compounds in JC tend to be “semantically

transparent”, composed of a wh-morpheme and a questioned semantic unit (Patrick 2007), as

in wich-paat ‘where’ wa-mek ‘why’, wa-taim ‘when’, uu-fa ‘whose’, etc. These wh-elements

are placed in a sentence initial position, but like for Yes/No questions, there is no subject-

auxiliary/TMA inversion:


27

15) a. Wich-paat Jan behn fling di rakstuon?

Where John PAST fling DET stone

“Where did John throw the stone?”

b. *Wich-paat behn Jan fling di rakstuon?

Where PAST John fling DET stone

Additionally, the particle a may occur preceding a wh-element:

16) A wich-paat Jan behn fling di rakstuon?

A where John PAST fling DET stone

“Where did John throw the stone?”

This a-particle appears to be the same particle used in focalized constructions:

17) A di rakstuon mi a luk fa.

A DET stone 1SG PROG look for

“It’s the stone I am looking for.”

It must be noted however, that whereas this a-element obligatorily precedes the fronted-

focalized phrases, it is optional in the case of wh-phrases (Durrleman-Tame 2008; Durrleman

& Shlonsky 2015). Another phenomenon involving fronting in JC is topicalization. Both

adjuncts and arguments may be topicalized as in (18) and (19) respectively:

18) Tumaro mi wi ramp wid di pikni dem. (Durrleman-Tame 2008)

Tomorrow 1SG FUT play with the child 3PL

“Tomorrow I will play with the children.”

19) Da bwai de, mi laik im. (Durrleman-Tame 2008)

That boy LOC 1SG like him

“As for that boy, I like him.”


28

Unlike topicalization of an adjunct, topicalization of an argument requires the presence of a

pronominal element in the comment to refer back to it (Durrleman-Tame 2008).

Nominals are not morphologically marked for nominative and accusative case, or for gender.

For genitive, they are also unaltered, although they are sometimes preceded by a prepositional

element fi, as in fi mi ‘my’ fi yu ‘your’, fi dem ‘their’, etc.

Based on the description of the language situation given above, strict methodological

considerations must be made in terms of selecting the right participants, whose linguistic

repertoire contains the typical creole features that are found at the basilectal end of the

continuum. This selection process will be discussed in the following chapter.

Chapter 4: Methodology

29

CHAPTER 4

METHODOLOGY

4.1 INTRODUCTION

This dissertation relies on naturalistic speech data compiled from a new corpus of longitudinal

development of the early stages of Jamaican Creole (JC). In this chapter, I outline the

methodological background of this dissertation. In section 4.2, I describe the participants,

their environments and the selection criteria. In section 4.3, I describe the data collection

procedures and the recording. And in Section 4.4, I outline the procedures of transcribing,

coding and analyzing the data.

4.2. PARTICIPANTS

In order to investigate the emergence of the early syntactic systems of JC, six children, age

ranging from 18 to 23 months, were studied for a period of 18 months. This age-range

corresponds to the period in which syntax typically emerges in children and during which

target-inconsistent forms and structures have been documented in other languages. The

linguistic production of children younger than 18 months is often too poor and too dispersed

to provide coherent data, while around 36-40 months, children are ‘talking machines’ and data

collection based on naturalistic recordings is generally much less informative than a

methodology based on production experiments. Experimental studies however tend to be

relatively narrow in scope and so cannot offer a broad perspective on children’s overall

linguistic development, to which this study is aimed. Being that the acquisition of JC is an

understudied area, a longitudinal corpus study is an excellent starting point providing a broad


30

overview of the relevant phenomena, which can be later refined through targeted experimental

work.

Subjects were strategically chosen from households where JC was predominantly spoken, and

as such there was less interference from English in the child’s linguistic environment. Given

the existence of the creole continuum, various factors were considered in identifying and

selecting the participants for inclusion in the study. Primary consideration was given to the

area of residence and the level of education of the primary caregiver. More specifically,

speakers from rural communities with less education were ranked closer to the basilectal end

of the continuum (Meade, 2001). In light of this observation, in the search for children to be

included in the study, I targeted the parish of St. Elizabeth. The area selected included the

adjoining communities of Southfield, Back Flagaman and Roundhill. This area was selected

since it was likely to have speakers who primarily used a basilectal variety of JC. This choice

was thus based largely on the socio-demographic profile of the residents, and general opinions

on where the most conservative JC could be found. Statistics for this area5 reveal that 74% of

household heads have no educational certification and there are no secondary or tertiary

educational facilities in these communities. 31% of the population are unemployed, while the

majority of the employed were predominantly in farming and fishing, with only 4% being

trained professionals (Francis, 2012).

In order to find participants in the desired age group, I consulted the Southfield Health Centre,

the only health facility in the area where all children of the community and surrounding areas

are expected to be registered. A letter was sent to the Head Nurse, explaining the research

5 These are statistics for the general Pedro Plains area, which includes Flagaman and Roundhill, but for which

Southfield is an adjoining community.


31

objectives and the rationale behind seeking participants. A list of prospective participants and

their contact details was received.

I arranged meetings with the parents and conducted preliminary interviews with caregivers of

the prospective participants. This study was guided by ethical principles and as such the aims

and objectives of the project, together with details about the data collection process were duly

explained to the participants. It was mandatory that the parents of all the participants signed

an informed consent form (see Appendix 1).

During the interviews, I was mindful of the observers’ paradox (Labov, 1972). This is where a

researcher tries to observe naturalistic speech, however his/her presence as an observer creates

a situation in which speakers are highly conscious of their speech and will therefore modify it.

In order to minimize speech modification, the interviews were informally conducted in the

homes of the prospective participants. For as much as possible, the language used for all

interactions with the prospective participants was JC. This thereby minimizes the likelihood

of influence from English, as these preliminary interviews allowed us to analyze the language

used by the caregivers and members of the household for typical creole features (as described

in Chapter 3, and in Bailey 1966, Patrick 2004, Durrleman-Tame 2008, etc.). If these features

were sufficiently present, then children in such households would be eligible subjects for

participation in the research project. Notwithstanding this, further selectional criteria were

applied, involving the willingness and availability of the informants and the level of speech

production by the children.

At the very initial interview, one participant was ruled out, as despite their overwhelming

interest in the research, the language of the household was not fitting with the basilectal


32

variety which the research targeted. After 3 sessions of recording, another informant was still

not producing any words despite his mother’s initial report that he was already speaking a few

words. To continue recording him on the basis of faith that he would speak in the next

recording session was not profitable for the research project; he was thus subsequently

replaced by another informant. A third informant was clearly not interested in participating in

the research project and was also removed. Additional informants were included in the study

based on references received from other participants. At the end of the selection process the

following table represents the final pool of participants included in the study.

Code Age at 1st recording

6 Gender Location

COL 1;6,11 male Back Flagaman

ALA 1;7,19 female Southfield

RJU 1;7,28 male Back Flagaman

TYA 1;9,18 female Round Hill

KEM 1;11,3 male Round Hill

SHU 1;11,25 female Back Flagaman

Table 1: Research Participants

Apart from the socioeconomic background, age and gender of the participants, various other

social variables could be expected to influence a child’s language development, such as size

of family, position in the birth order, presence of both or only one parent, attendance at

nursery, etc. (Wells, 1985). These were allowed to vary randomly, as it would be difficult to

keep them constant throughout the entire duration of the study.

6 Age is presented in Year; Month, Day format.


33

I present in the next section individual profiles for all participants.

4.2.1 Participants’ Individual Profiles

COL (age range: 1;6,11 – 2;11,7): COL is the youngest informant in the study. He had a

Mean Length of Utterance (MLU) of 1.6 at 20 months and 3.81 at 35 months. He was a single

child living with both parents in the community of Back Flagaman. They lived in a ‘nestled’

area where five other houses were a stone throw away. These houses contained other children,

one of whom was also a participant of this study. In addition, his paternal grandmother and

cousins were immediate neighbours, and as such the yard was normally active. His maternal

grandmother lived in the same community, before migrating abroad, 9 months into the study.

Some of his recordings were conducted at her house; however the majority was done in his

home. His father was a farmer, who habitually fished also, while his mother was a housewife.

After starting kindergarten at 28 months old, his parents constructed a small shop in their

yard, where the mother worked as a shopkeeper.

ALA (age range: 1;7,19 – 3;0,15): At 21 months, with an MLU value at 1.48 and 5.66 at 36

months, she was one of the most vocal participants in the study. She lived in the community

of Southfield with her parents in a family house where they occupied a room at the back. She

was a single child for her parents but had cousins, aunts, uncles and grandparents in the

extended household. Her mother held a clerical/administrative position in a governmental

agency in the parish, while her father was a driver distributing goods for a furniture company.

As such, when both parents were at work, the child would stay with relatives in the

nieghbouring community of Seaview. Recordings were therefore conducted at both Seaview

and at the child’s home. ALA started attending school at 33 months old.


34

RJU (age range: 1;7,28 – 3;0,25): RJU lived in an extended family household with his

parents, paternal grandmother, aunts, uncles, and cousin. His cousin, a girl aged 6 years,

appeared to be his best friend with whom he interacted the most. His mother was unemployed

while his father worked as a farmer. Throughout the 18 month period, RJU visited his paternal

grandfather in a district located about 10 miles away from his home community. Some

recordings were conducted there. For the last four months of the recording sessions, his

mother separated from his father and lived with other relatives in the neighboring community

of Crossroads. Some of his recordings were conducted there. RJU later returned home under

the care of his paternal grandmother and aunt. The majority of his recordings were conducted

in his home in the presence of his cousin. At 22 months, he had an MLU of 1.39 and ending

with an MLU of 4.86 at 36 months.

TYA (age range: 1;9,18 – 3;2,15): TYA lived in the community of Roundhill with her parents

and two siblings. Her mother was a housewife while her father was a taxi-driver. Her

immediate neighbours were her grandmother, aunts and cousins. She had an MLU of 1.22 at

23 month and 4.86 at 38 months. She started school at 35 months; at this point we saw a very

rapid development in her speech, moving from an MLU of 2.16 at 34 months of age to a high

of 5.38 at 37 months. Her recordings were conducted mainly in the comfort of her home.

KEM (age range: 1;11,3 – 3;3,11): KEM’s MLU was 2.1 and 5.47 at 24 and 39 months

respectively, peaking at 6.46 at 37 months. He started school at 32 months. He lived with his

mother, maternal grandparents, aunt and uncle in the community of Roundhill. His mother

was unemployed, his grandfather was a farmer and his grandmother operated a shop in the

yard. The shop was frequented by elderly farmers, some of whom were alcoholics, using quite

“colourful” language. This linguistic behavior was quickly reflected in the child’s speech.


35

KEM sometimes had playmates with whom he ran about freely in his large yard space and

neighbouring taverns. All of his recordings were conducted at his home.

SHU (age range: 1;11,25 – 3;4,13): SHU is the eldest participant in the research project. She

had an MLU of 2.88 at 25 months and 5.02 at 40 months. She lived in the community of Back

Flagaman with her mother and her brother and enjoyed a visiting relationship with her father

and his family who lived just a few meters away. She was the only child for her father. Upon

the passing of her father at 28 months, she lived with her paternal grandparents, aunt and

uncle, and then had a visiting relationship with her mother. Her mother was generally

unemployed but worked occasionally as a store clerk. Her grandfather operated a shop and her

grandmother was a housewife. She started school at 33 months of age. Recordings were

conducted mainly at the home of her mother or paternal grandparents, and on a few occasions

at the home of her maternal grandmother.

MLU values are plotted in Figures 1 – 6 showing a steady increase of utterance length against

age for all participants in the study. On the x-axis we present the MLU values and on the y-

axis the informants’ age in year; months, days format. The complete list of MLU values are

given in Tables (2) – (7). Note that this does not include the 2 initial months of data collection

and only two recordings per month are represented. Justification for this is presented in

section 4.4. The MLUs were calculated automatically with the help of CLAN (Computerized

Language Analysis). They are all word based, as opposed to morpheme based. Being an

isolating language, all lexical and functional elements are counted as independent words. The

MLU presented for JC may therefore not be immediately comparable to that in languages

with morphologically complex words.


36

Figure 1: COL’s MLU Figure 2: ALA’s MLU

Figure 3: RJU’s MLU Figure 4: TYA’s MLU

1

1.5

2

2.5

3

3.5

4

4.5

1;8,

17

1;9,

28

1;11

,12

2;0,

28

2;2,

16

2;3,

30

2;5,

14

2;6,

25

2;8,

6

2;9,

24

2;11

,71

1.5

2

2.5

3

3.5

4

4.5

5

5.5

6

1;9,

25

1;11

,5

2;0,

20

2;2,

6

2;3,

24

2;5,

7

2;6,

22

2;8,

2

2;9,

14

2;11

,1

3;0,

15

1

1.5

2

2.5

3

3.5

4

4.5

5

5.5

1;1

0,4

1;1

1,1

5

2;0

,30

2;2,

16

2;4

,4

2;5

,17

2;7

,5

2;8

,12

2;9,

24

2;1

1,1

1

3;0

,25

1

1.5

2

2.5

3

3.5

4

4.5

5

5.5

6

1;1

1,2

5

2;1

,5

2;2

,20

2;4,

6

2;5

,24

2;7

,7

2;8

,22

2;1

0,2

2;11

,14

3;1

,1

3;2

,15


37

Figure 5: KEM’s MLU Figure 6: SHU’s MLU

Table 2: COL’s MLU Table 3: ALA’s MLU

1

2

3

4

5

6

7

2;0,

21

2;2,

1

2;3,

16

2;5,

2

2;6,

20

2;8,

3

2;9,

18

2;10

,29

3;0,

10

3;1,

28

3;3,

11

1

1.5

2

2.5

3

3.5

4

4.5

5

5.5

6

2;1,

23

2;3,

3

2;4,

18

2;6,

4

2;7,

22

2;9,

5

2;10

,20

3;0,

0

3;1,

12

3;2,

30

3;4,

13

RECDATE AGE(Y;M,D) UTT WORDS MLU

16.01.2012 1;8,17 160 256 1.6

26.01.2012 1;8,27 181 325 1.796

16.02.2012 1;9,17 219 392 1.79

27.02.2012 1;9,28 191 287 1.503

09.03.2012 1;10,8 261 401 1.536

31.03.2012 1;11,1 251 449 1.789

11.04.2012 1;11,12 194 389 2.005

27.04.2012 1;11,28 230 421 1.83

12.05.2012 2;0,12 138 343 2.486

28.05.2012 2;0,28 162 410 2.531

13.06.2012 2;1,14 240 497 2.071

30.06.2012 2;2,0 127 279 2.197

16.07.2012 2;2,16 354 813 2.297

31.07.2012 2;3,1 196 503 2.566

15.08.2012 2;3,16 215 615 2.86

29.08.2012 2;3,30 239 678 2.837

14.09.2012 2;4,15 238 679 2.853

30.09.2012 2;5,0 200 551 2.755

14.10.2012 2;5,14 193 606 3.14

27.10.2012 2;5,27 401 1110 2.768

09.11.2012 2;6,10 218 555 2.546

24.11.2012 2;6,25 413 1407 3.407

08.12.2012 2;7,8 341 1190 3.49

22.12.2012 2;7,22 266 866 3.256

05.01.2013 2;8,6 339 1235 3.643

19.01.2013 2;8,20 343 1429 4.166

10.02.2013 2;9,11 163 611 3.748

23.02.2013 2;9,24 271 907 3.347

12.03.2013 2;10,10 275 1042 3.789

23.03.2013 2;10,21 271 933 3.443

06.04.2013 2;11,7 267 1019 3.816

COL


16.01.2012 1;9,25 271 401 1.48

26.01.2012 1;10,4 208 254 1.221

16.02.2012 1;10,25 385 634 1.647

27.02.2012 1;11,5 349 551 1.579

09.03.2012 1;11,16 341 632 1.853

31.03.2012 2;0,9 381 856 2.247

11.04.2012 2;0,20 254 458 1.803

27.04.2012 2;1,5 395 794 2.01

12.05.2012 2;1,20 353 852 2.414

28.05.2012 2;2,6 377 1012 2.684

13.06.2012 2;2,22 370 1140 3.081

30.06.2012 2;3,8 117 291 2.487

16.07.2012 2;3,24 397 1266 3.189

31.07.2012 2;4,9 283 933 3.297

15.08.2012 2;4,24 388 1237 3.188

29.08.2012 2;5,7 294 1025 3.486

14.09.2012 2;5,23 493 2111 4.282

04.10.2012 2;6,12 193 623 3.228

14.10.2012 2;6,22 227 937 4.128

27.10.2012 2;7,5 336 1382 4.113

09.11.2012 2;7,18 361 1448 4.011

24.11.2012 2;8,2 467 2547 5.454

08.12.2012 2;8,16 313 1729 5.524

22.12.2012 2;9,0 294 1251 4.255

05.01.2013 2;9,14 285 1336 4.688

19.01.2013 2;9,28 249 1259 5.056

16.02.2013 2;10,25 297 1351 4.549

23.02.2013 2;11,1 246 1078 4.382

12.03.2013 2;11,18 264 1292 4.894

23.03.2013 3;0,1 277 1353 4.884

06.04.2013 3;0,15 341 1930 5.66

ALA


38

Table 4: RJU’s MLU Table 5: TYA’s MLU


16.01.2012 1;10,4 51 71 1.392

26.01.2012 1;10,14 80 137 1.712

16.02.2012 1;11,4 198 440 2.222

27.02.2012 1;11,15 39 65 1.667

09.03.2012 1;11,26 66 105 1.591

31.03.2012 2;0,19 135 268 1.985

11.04.2012 2;0,30 147 332 2.259

27.04.2012 2;1,15 207 692 3.343

12.05.2012 2;2,0 202 528 2.614

28.05.2012 2;2,16 173 361 2.087

13.06.2012 2;3,1 315 950 3.016

30.06.2012 2;3,18 206 604 2.932

16.07.2012 2;4,4 207 605 2.923

31.07.2012 2;4,19 212 674 3.179

15.08.2012 2;5,3 136 435 3.199

29.08.2012 2;5,17 220 570 2.591

14.09.2012 2;6,2 206 720 3.495

30.09.2012 2;6,18 268 762 2.843

17.10.2012 2;7,5 249 1071 4.301

27.10.2012 2;7,15 257 1031 4.012

09.11.2012 2;7,28 209 919 4.397

24.11.2012 2;8,12 276 1040 3.768

08.12.2012 2;8,26 233 859 3.687

22.12.2012 2;9,10 236 773 3.275

05.01.2013 2;9,24 226 780 3.451

19.01.2013 2;10,7 187 730 3.904

10.02.2013 2;10,29 142 452 3.183

23.02.2013 2;11,11 173 626 3.618

09.03.2013 2;11,25 328 1662 5.067

23.03.2013 3;0,11 206 1079 5.238

06.04.2013 3;0,25 277 1346 4.859

RJU


16.01.2012 1;11,25 126 154 1.222

26.01.2012 2;0,4 130 157 1.208

16.02.2012 2;0,25 113 153 1.354

27.02.2012 2;1,5 257 324 1.261

09.03.2012 2;1,16 108 152 1.407

31.03.2012 2;2,9 121 162 1.339

11.04.2012 2;2,20 174 246 1.414

27.04.2012 2;3,5 101 149 1.475

12.05.2012 2;3,20 218 334 1.532

28.05.2012 2;4,6 162 235 1.451

13.06.2012 2;4,22 128 169 1.32

30.06.2012 2;5,8 62 99 1.597

16.07.2012 2;5,24 103 152 1.476

31.07.2012 2;6,9 58 111 1.914

15.08.2012 2;6,24 205 411 2.005

29.08.2012 2;7,7 22 52 2.364

14.09.2012 2;7,23 240 467 1.946

30.09.2012 2;8,8 302 926 3.066

14.10.2012 2;8,22 143 401 2.804

27.10.2012 2;9,5 233 696 2.987

09.11.2012 2;9,18 53 99 1.868

24.11.2012 2;10,2 260 635 2.442

08.12.2012 2;10,16 178 385 2.163

22.12.2012 2;11,0 137 510 3.723

05.01.2013 2;11,14 271 914 3.373

19.01.2013 2;11,28 261 1168 4.475

10.02.2013 3;0,19 42 194 4.619

23.02.2013 3;1,1 268 1442 5.381

09.03.2013 3;1,15 325 1722 5.298

23.03.2013 3;2,1 177 923 5.215

06.04.2013 3;2,15 175 851 4.863

TYA


39

Table 6: KEM’s MLU Table 7: SHU’s MLU

4.3 RECORDINGS

Preliminary interviews and recordings were conducted at the end of October 2011 and

recording sessions got on the way in November 2011. Dr. Durrleman visited in mid-

November and took part in the recording sessions and the decisions regarding the final

selection of participants. A graduate student from the University of the West Indies, Nickesha


16.01.2012 2;0,21 280 589 2.104

26.01.2012 2;1,0 348 654 1.879

16.02.2012 2;1,21 309 552 1.786

27.02.2012 2;2,1 297 574 1.933

09.03.2012 2;2,12 423 821 1.941

31.03.2012 2;3,5 305 600 1.967

11.04.2012 2;3,16 347 725 2.089

27.04.2012 2;4,1 392 933 2.38

12.05.2012 2;4,16 337 767 2.276

28.05.2012 2;5,2 279 753 2.699

13.06.2012 2;5,18 386 929 2.407

30.06.2012 2;6,4 303 882 2.911

16.07.2012 2;6,20 376 1148 3.053

31.07.2012 2;7,5 373 1513 4.056

15.08.2012 2;7,20 311 1322 4.251

29.08.2012 2;8,3 258 1103 4.275

14.09.2012 2;8,19 276 1287 4.663

30.09.2012 2;9,4 341 1607 4.713

14.10.2012 2;9,18 261 1170 4.483

27.10.2012 2;10,1 159 683 4.296

09.11.2012 2;10,14 254 1141 4.492

24.11.2012 2;10,29 404 1884 4.663

08.12.2012 2;11,12 265 1160 4.377

22.12.2012 2;11,26 355 2203 6.206

05.01.2013 3;0,10 285 1606 5.635

19.01.2013 3;0,24 333 2151 6.459

10.02.2013 3;1,15 261 1640 6.284

23.02.2013 3;1,28 339 1937 5.714

09.03.2013 3;2,11 384 2397 6.242

23.03.2013 3;2,25 384 2278 5.932

06.04.2013 3;3,11 295 1614 5.471

KEM


16.01.2012 2;1,23 177 511 2.887

26.01.2012 2;2,2 233 705 3.026

04.02.2012 2;2,11 191 517 2.707

27.02.2012 2;3,3 307 1059 3.45

09.03.2012 2;3,14 287 1147 3.997

31.03.2012 2;4,7 91 287 3.154

11.04.2012 2;4,18 133 368 2.767

27.04.2012 2;5,3 175 604 3.451

12.05.2012 2;5,18 372 1084 2.914

28.05.2012 2;6,4 239 893 3.736

13.06.2012 2;6,20 234 755 3.226

10.07.2012 2;7,16 332 1157 3.485

16.07.2012 2;7,22 244 894 3.664

31.07.2012 2;8,7 244 1080 4.426

15.08.2012 2;8,22 171 663 3.877

29.08.2012 2;9,5 281 921 3.278

14.09.2012 2;9,21 333 1387 4.165

30.09.2012 2;10,6 195 695 3.564

14.10.2012 2;10,20 58 198 3.414

27.10.2012 2;11,3 284 1001 3.525

09.11.2012 2;11,16 156 524 3.359

24.11.2012 3;0,0 299 1340 4.482

13.12.2012 3;0,19 283 1440 5.088

22.12.2012 3;0,28 375 1900 5.067

05.01.2013 3;1,12 285 1366 4.793

19.01.2013 3;1,26 509 2883 5.664

10.02.2013 3;2,17 193 942 4.881

23.02.2013 3;2,30 226 1122 4.965

09.03.2013 3;3,13 191 623 3.262

23.03.2013 3;3,27 306 1626 5.314

06.04.2013 3;4,13 331 1661 5.018

SHU


40

Dawkins, was employed as a Research Assistant to the project in December 2011, who

assisted in carrying out the recordings and transcriptions of the data.

The corpus was collected by making digital recordings of the children’s spontaneous speech

in a natural setting. Each child was recorded every 10 days in separate 60 minutes sessions,

for the initial five months, and subsequently approximately every 15 days for the remaining

duration of the data collection phase of the project. This recording method proved to be

advantageous as it yielded a large amount of data covering a wide range of linguistic

phenomena, allowing a comprehensive picture of the participants’ overall development. Not

accounting for the first two months when selection and familiarization processes were still

underway, we ended-up with a database of 204 hours of recordings. This provides a relatively

large database of early child JC utterances. Recordings were conducted in JC, in the presence

of the interviewers, care-providers and in some instance other relatives and friends. The

format of the recording sessions was informal and participants, becoming close friends with

the researchers, spoke freely on various subject matters. As the aim of the sessions was to

obtain naturally occurring speech, the informants were allowed to move freely around their

home and in its immediate vicinity, interacting with others in a normal manner. The sessions

were usually interactively built upon games and storybooks. The researchers would take a

small gift for each child at every recording session to maintain motivation.

Based on the close proximities of the communities in which the children resided, recording

sessions were usually all conducted in one day. The researchers would get in the field around

9 a.m. (sometimes earlier depending on the availability of the participants) and recordings

would begin immediately. Both researchers would interview ALA, as she was the only

participant of her community, located approximately 15 minutes by car from the communities


41

where the other informants reside. TYA and KEM lived in the same community and as such

the researchers conducted parallel recording sessions. Parallel recording sessions were usually

conducted with SHU and COL, and a joint session for RJU. This schedule was not always

maintained, as the sessions were flexible depending on the availability of the participants. On

occasions, where informants were unavailable on the scheduled recording day, another visit

would be arranged where the recording session would be conducted. As some participants

started school during the course of the research project, recordings were subsequently

scheduled to take place on the weekends, and in some instances after school.

Field visits however, did not proceed without their fair share of distractors, ranging from

dysfunctional transportation means of reaching the research site to being greeted by disruptive

outsiders intruding on the recording sessions. Nonetheless, the opportunity gained by

becoming involved in the lives of the informants and their households in collecting most

valuable data outweighed all.

Hand held digital voice recorders were used as the main tool for data collection. Initial

recordings were conducted with the recorder attached to the child bearing a mike, however

this proved to be problematic as, not only was it a major distractor for the child, but the use of

a single input mike allowed only for the audible processing of the child’s data and not the

surrounding interlocutors. Where necessary, notes were recorded after the sessions.

4.4 TRANSCRIPTION, CODING AND ANALYSIS

JC is mainly an oral language. Many of the lexemes are English based but their phonology is

quite different. For all transcriptions and examples reproduced in this dissertation, the JLU


42

(Jamaica Language Unit) modified Cassidy-LePage orthography was employed. This is a

system that represents the sounds of JC as faithfully as possible, without relying on the

spelling conventions of English. This system has no silent letters and there is a one-to-one

mapping of sound to symbol therefore each letter or letter combination is always pronounced

the same way (Jumieka Langwij).

All data was transcribed in CHAT (Codes for the Human Analysis of Transcripts) format,

following the standard guidelines of the CHILDES (Child Language Data Exchange System)

Database. However, for clarity, most of the codes have not been used. The transcription

procedure proved to be very time consuming as it took approximately 10 hours to manually

transcribe one hour of data. For transcribing the data, the recordings were transferred from the

recording device to the computer. High quality earphones were used and the researchers

transcribe the exact production as uttered. In most instances, due to surrounding noise

including overlapping speech, recordings had to be repeatedly listened to in order to ensure

accurate transcription. Inaudible speech was transcribed as xxx.

While the joint recording sessions were transcribed by the main researcher, the recordings of a

particular child in the parallel sessions were transcribed by the researcher who had conducted

the recordings. In this way all the paralinguistic details not included in the audio recordings,

but which could have impacted the interpretation of a particular utterance could be included in

the transcriptions.

A backed-up copy of all recordings and transcriptions is stored on the University of Geneva

Database for safekeeping. Moreover, the database will be transferred to CHILDES for public

access.


43

4.4.1 Coding

Coding entails analyzing the transcriptions and making notes of the grammatical and syntatic

phenomena. This is done by creating a one-to-one correspondence between utterances and

standard morpho-syntatic codes on a word-by-word basis. Based on the time frame for the

completion of the research project, a decision was taken to initially code only the utterances

of the target children. I started coding the first set of transcriptions, however, it proved

unfeasible and as such four final-year students from the University of the West Indies were

subsequently employed to complete the coding of the data, under my guidance. The

approximate time to code one transcription was 6 hours. This time however is not fixed as it

depends largely on the number and length of child utterances in the transcription.

In order to select competent students for employment for this task, a class of Creole

Linguistics final year students was trained in doing transcriptions and coding data in the

CHAT format. The four most outstanding students in this task were subsequently employed to

undertake the morpho-syntactic coding of the corpus.

Based on the non-standard conventions in transcribing JC, coding of the data had to be done

manually. A list of codes was developed for conveying the morpho-syntactic relations (See

Appendix 2). Despite this comprehensive list, coding of the data did not prove to be

unproblematic as there are some instances where a particular lexical item could lead to

different interpretations or yield different codes in the same context. To deal with these

occurrences, native speakers’ judgments were employed where applicable, or the word in

question coded as unknown.


44

In dealing with other issues regarding the coding of utterances, we adopted the method

employed in comparative syntax, assuming on grounds of continuity and uniformity, that

child language approximates adult grammar (in line with Bates et.al. 1994, Gleitman et.al.

1999, among others). We acknowledge that this is not necessarily true in general but it is the

necessary initial assumption for comparative research. The data was also coded for null

elements in the grammar.

Several meetings were conducted with the coding team to resolve all issues. Two recordings

were coded per child for the period starting January 2012 to March 2013 and one for each

child in April 2013. The decision to start coding the data for analysis as of January 2012 and

to include only two of the three recording sessions for the period January to March 2012 was

mainly based on the following:

1. the maximal use of funds available

2. standardizing the quantificational dimension of the analysis

3. the final selection and confirmation of all the research participants

4. the initial two months involved familiarization of the participants with the researchers

thereby maximizing their language production levels.

All completed coding was duly checked for verification of accurateness, and for inclusion of

additional details as required for the analysis. A total of 186 transcripts, comprising more than

80,000 meaningful child utterances (not including fillers e.g. mh) were coded and

subsequently became the basis of analysis for this dissertation.


45

4.4.2 Analysis

The coding scheme provides for the description and analysis of the data in terms of syntactic

structures. The analysis of the production data was based mainly on age and developmental

stages (in line with Radford 1990). Nonetheless, where necessary the participants’ production

was classified and compared by their mean length of utterance (MLU) as this method is

widely considered to measure the level of language development (see Brown 1973, Miller

1981, Miller and Chapman 1981). Since JC is an isolating language, lacking morphologically

complex words, MLU was calculated on a word basis. Single word utterances of fillers such

as mmh were not included in the count.

Some utterances were excluded from the data analysis, these include:

- utterances in which any unintelligible portions (coded an UNK) could be critical for the

analysis

- utterances where the meaning was unclear based on the context of the discourse

- the child’s stuttering or self-repetitions without the production of contentful utterances

in-between

- repetitions of memorized materials, e.g. songs and nursery rhymes

- immediate repetitions of adult’s exact utterance

Various sections of the analysis are based on automatic computing of the morpho-syntactic

coding using, but not limited to, CLAN, NotePad++ and Excel. Nonetheless, manual analysis

was inevitable for certain computations. Of course, in order to conduct cumulative analysis of

the group as a whole, and for developmental comparisons, the exact age of each child could

not be used. I therefore grouped the participants by approximate age. For example, a child


46

who was aged 2 years, 6 months and 4 days (2;6,4) was categorized as 2 years and 6 months

old (2;6.0) and one who was 2 years, 6 months and 20 days (2;6,20) categorized as 2 years,

6.5 months old (2;6.5).

The analysis provided in this dissertation is generally wide-scoped, covering a range of

syntactic phenomena in a comparative manner. Given our present knowledge of the target

system, as alluded to in the previous chapter, the syntax of adult JC is relatively well

documented, thereby facilitating comparisons between the child system and the adult

grammar. Additionally, the enormous published theoretical and descriptive work in non-

creole child language acquisition enables comparative analyses among the syntactic systems

of children in various languages. The analysis of the data therefore seeks to explain not only

the comprehensive development of early JC, but to add to our understanding of the general

nature of language acquisition.

Chapter 5: General Developmental Patterns in JC

47

CHAPTER 5

GENERAL DEVELOPMENTAL PATTERNS IN JC

5.1 INTRODUCTION

The goal of this chapter is to provide a step by step descriptive view of the general

development patterns of clauses in early Jamaican Creole (JC) and to propose an analysis of

phrase structure in the initial stages of acquisition. I provide details of the developmental

stages and the changes in acquisition as children move towards the target system, outlining

whether the developmental patterns reflect gradual changes or whether there is immediate

categorical eradication of the previous grammatical system. Section 5.2 outlines the system of

grouping which thereby enables the comparative analysis of early grammatical development.

In section 5.3 I examine the types of utterances produced by early learners of JC and the

general production of grammatical categories. In section 5.4 I describe the two word stage,

examining the initial combination of words and the basic principles of phrase and sentence

formation. Section 5.5 explores the development of multiword utterances, examining in

particular the expansion of the Determiner Phrase, the Inflectional Phrase and the

Complementizer Phase. Section 5.6 provides a discussion and summary of the main

conclusions.

5.2 DEVELOPMENTAL STAGES

Based on the large number of individual items involved across the participants, there is an

absolute need for cumulative grouping across developmental stages. One traditional view


48

(Stern 1928, cited in Radford 1990) is to divide early grammatical development according to

the following stages:

0-12 months: prelinguistic stage (babbling stage before development of words)

12-18 months: single-word stage (utterances comprise a single word spoken in isolation)

18-24 months: early multi-word stage (combination of 2, 3 or 4 words to form productive

syntactic structures)

24-30 months: later multi-word stage (structures with 5 or more words become

productive)

Radford (1990) presented a system where development is grouped categorically as follows:

Precategorical stage (one word speech when categorization has not yet taken place)

Lexical stage: around the age of 20 months (acquisition of a set of lexical categories

and their phrasal projections)

Functional stage: around 24 months (acquisition of functional word categories and

their phrasal projections)

As children acquire language at widely varying rates, instead of grouping by age, Brown

(1973) grouped children according to their mean length of utterance (MLU). According to

Brown (ibid.) the MLU is an excellent simple index of grammatical development because

almost every new kind of knowledge increases length. He proposed that matching for MLU is

more likely to produce similar speech patterns in terms of constructional complexity than

matching chronological age. He therefore provides the following grouping:


49

Stage MLU

I 1.75

II 2.25

III 2.75

IV 3.50

V 4.00

In examining the present data, it is seen where there is a vast difference in terms of age and

level of production. For example, while most children at 27 months would be producing 3

word combinations, TYA is still producing single word utterances. Despite the fact that all the

children in our study had the same period of time and relatively unrestricted opportunities to

speak, there is considerable variation in both the quantitative and qualitative speech (in terms

of grammatical developments) that was actually produced when comparisons are based on

groupings by age. Therefore in handling the current data it will not be profitable to adopt a

grouping based on age in line with Stern’s (1928) or Radford’s (1990) categorization.

Nonetheless it is possible to make direct comparisons between the children with respect to

their overall production and development of speech in line with their MLU. Being that JC is

an analytic language, a categorization by MLU calculated on a word level would accurately

map the production of children in terms of development and combination of words. Brown’s

categorizations however were very arbitrarily designed to account specifically for his dataset.

They are not based on qualitative changes reflected in the data itself, but were rather a simple

device for sampling the large quantity of data. For the current research, I believe a less

arbitrary approach to dividing the data would be more profitable. As MLU is calculated on a

word level, I will map development in terms of word/morpheme combinations. I will

therefore propose the following categorization of speech development for JC.


50

MLU <1.49 Single word utterances

MLU 1.50 – 2.49 First combinations (2 words)

MLU 2.5 – 3.49 3-words combinations

MLU 3.5 – 4.49 4-words combinations

MLU 4.5 > Complex combinations (5 or more words)

The age range which corresponds to these categories for each child is presented in Table 1. As

can be seen, the ages in which these developments are attested vary. Nonetheless, I will

attempt to describe the general pattern of MLU development in relation to age as attested in

the present corpus. The single word stage corresponds roughly to an age under 22 months for

ALA and RJU but up to 27 months for TYA. At the start of the research COL, KEM and SHU

were already beyond the single word stage. The 2-word combination stage goes up to an age

ranging from 24 months to 28 months, but in the case of TYA, this stage goes up to 31

months. SHU’s production was still more advanced at the start of her recordings. The end of

3-words combination corresponds to the age ranging from 29 to 31 months of age, but up to

35 months for TYA. After this age, TYA’s development appears to be now on par with the

other children in the corpus. The combination of 4-words goes up to an age ranging from 31

to 36 months. This stage represents the last production in COL’s dataset, however all other

informants produced combinations of 5 or more words after this age.


51

MLU COL ALA RJU TYA KEM SHU

<1.49 - 1;9,25 – 1;10,4 1;10,4 1;11,25 – 2;3,5 - -

1.50 – 2.49 1;8,17 – 2;0,12 1;10,25 – 2;1,20 1;10,14 – 2;0,30 2;3,20 – 2;7,23 2;0,21 – 2;4,16 -

2.5 – 3.49 2;0,20 – 2;7,22 2;2,6 – 2;5,7 2;1,15 – 2;6,18 2;8,8 – 2;10,16 2;5,2 – 2;6,20 1;11,25 – 2;7,16

3.5 – 4.49 2;6,6 – 2;11,7 2;5,23 – 2;7,18 2;7,5 – 2;11,11 2;11,0 – 2;11,28 2;7,5 – 2;8,3 2;7,22 – 3;0,0

4.5 > - 2;8,2 – 3;0,15 2;11,25 – 3;0,25 3;0,9 – 3;2,15 2;8-19 – 3;3,11 3;0,19 – 3;4,13

Table 1: Participants age range in relation to MLU

Based on these developmental categories, we see that the single word, early combinations and

later multiword utterances do not correspond with the age groupings of Stern (1928); however

they map more directly with Radford’s lexical and functional stages. For example, they allow

us to say whether functional items are present in single word or 2-words utterances, or

whether they are only productively used when more complex combinations begin.

Additionally, this grouping will enable a clear investigation in the acquisition of grammar;

whether grammar emerges incrementally in line with the ‘structure building’ approach to

development (as supported by Radford (1990) for example), or whether both lexical and

functional structures are equally available from the start of significant linguistic development

in line with the ‘full competence’ approach to development (as advocated by Poeppel &

Wexler 1993).

5.3 GRAMMATICAL CATEGORIES

This section describes the general development of the grammatical categories produced in

utterances of early learners of JC and the emergence of syntactic systems. I will adopt

Radford’s (1990) categorical groupings assuming that there are 4 primary lexical categories:


52

nouns, verbs, prepositions and adjectives, with adverbs being analyzed as sub-classes of

adjectives. I will however add another category and analyze pronouns within the lexical

domain, separate and apart from nouns. The three main functional classes to be analyzed are:

determiners, inflections and complementizers. We will examine the number of utterance types

during each stage, grouped as declaratives, wh-interrogatives, yes-no interrogatives and

imperatives.

5.3.1 Utterance Types

As demonstrated in Figure 1 and corresponding Table 2, most of the sentences produced by

the children are declaratives. In the single word stage where MLU is less than 1.49, we note

that declaratives accounts for 94% (1519 utterances) of the children’s production, while other

utterances were minimal. One may question the categorization and subsequent calculations

during this stage, being that the initial productions are single words and possibly need not be

strictly described as declaratives. Early declaratives may not be true declaratives as yet, as

their opposites, i.e. interrogative and imperative forms are still missing. The child thus does

not acquire the interrogative in addition to the already existing declarative, but instead

acquires both at the very same time by the process of feature specification. Before this

process, single word utterances can be +/- declaratives or +/- interrogatives or neither of the

two (Hohenberger 2002). However, in line with Wells (1985) these initial utterances most

frequently realize the functions of ‘ostension’ (i.e. a process of showing by pointing) and as

such they can be accorded the status as declaratives. In addition, for JC, the context and the

associated intonation of utterances are major cues to the unambiguous interpretation of


53

utterance type7 and were therefore employed as defining factors for the categorization

decisions that were made in the process of coding the data.

At the next stage where 2 words combination begins (MLU 1.5 – 2.49), the data reveals an

increase in the use of imperatives up to roughly 10%. This production of imperatives remains

more or less constant throughout all subsequent stages of development. Yes/No and Wh

interrogatives are still minimally attested in the data, i.e. under 3%. However, as MLU

increased over 2.5, we note a gradual increase in the use of interrogatives, with wh-questions

being more dominant than yes/no questions.

Figure 1: Distribution of Utterance Types

7 The diagnostics for grouping utterance types may be limited as intonational patterns were not strategically

measured but are nonetheless a salient criteria in JC.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

< 1.49 1.5 - 2.49 2.5 - 3.49 3.5 - 4.49 4.5 >

Wh-Interrogatives

Yes/No Interrogatives

Imperatives

Declaratives

MLU (words)

%UTTERANCE TYPES


54

MLU (words)

Total utterances Declaratives Imperatives

Yes/No Interrogatives

Wh- Interrogatives

< 1.49 1615 1519 55 22 19

1.5 - 2.49 8870 7565 830 230 245

2.5 - 3.49 14255 10977 1438 712 1128

3.5 - 4.49 10085 7392 1055 484 1154

4.5 > 12941 9267 1638 864 1172

Table 2: Distribution of Utterance Types

Generally speaking, declarative sentence types emerge with greater frequency and possibly

before imperative and interrogative types. This may suggest that there are additional demands

on processing ability in the production of non-declarative utterances. The declarative-first

sequence has been reported for studies of English acquisition, with the subject and auxiliary

inversion requirement in question formation being the determining factor of the complexity

(Wells, 1985). Based on the fact that in JC, question formation does not require inversions

and all sentence types are similarly constructed (as demonstrated in 1 – 4), an alternate reason

for the attested pattern must be considered.

1) Paasta shot im yai. (Declarative) (SHU 2;7)

Pastor shut 3SG eye

‘Pastor shut his eyes.’

2) Shot yu yai! (Imperative) (ALA 2;05)

Shut 2SG eye

‘Shut your eyes!’

3) (Yo) shot it? (Yes/No Interrogative) (KEM 2;03)

2SG shut 3SG.

‘(Did you) shut it?


55

4) (Wa-mek) yo shot yo yai? (Wh-Interrogative) (ALA 2;05)

What-make 2SG shut 2SG eye

‘(Why did) you shut your eyes?

One possible explanation is to rely on semantic or pragmatic reasons. Children’s initial

utterances are mainly to express wants, to call for attention, or to draw attention to some

object or event (Wells 1985) for which a declarative utterance may suffice. On the assumption

that this would apply to children acquiring JC, I therefore speculate that the declarative-first

sequence found in early JC child grammar is not in relation to the syntactic complexity of the

clause but rather based on pragmatic/semantic reasons.

5.3.2. Word Classes

We noticed that regardless of utterance types, the earliest word class categories produced by

JC children are lexical. Lexical categories accounts for 99.6% of utterances during the single

word stage, as demonstrated in Figure 2 and corresponding Table 3.


56

Figure 2: Distribution of Word-Class Categories

MLU (words) Total word-class Lexical Functional

< 1.49 1623 1616 7

1.5 – 2.49 12821 12592 229

2.5 – 3.49 33099 30776 2323

3.5 – 4.49 29566 26681 2885

4.5 > 46445 40550 5895

Table 3: Distribution of Word-Class Categories

There is a gradual increase in the production of functional items however; reaching a high of

12.7% during the stage described as complex combinations where MLU is over 4.5. The

predominantly lexical developmental pattern attested in the 2 initial stages is expected in line

with Radford (1990) where he proposed that early child grammars of English are purely

lexical structures. He argued that around the age of 20-23 months, children develop an N-

system, V-system, P-system and A-system. During this period, children have no D-system, C-

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

< 1.49 1.5 – 2.49 2.5 – 3.49 3.5 – 4.49 4.5 >

Functional

Lexical

MLU (in words)

% W

OR

D C

LASS

PR

OD

UC

TIO

N


57

system or I-system in their grammars. His explanation for this developmental order is that the

linguistic properties (morphophonosyntactic and semantic properties) of items belonging to

functional categories make them more difficult to learn and as such they are acquired later

than their lexical counterparts.

On the other hand, it can be argued that children have functional categories from the start,

even when there is no overt material corresponding to a particular projection (Rasetti 2003).

Though not robustly attested, the fact that 7 utterances with functional attributes were found

in the data in the single word stage and 229 in the initial combinations of 2 words could

provide evidence of a functional system. Additionally, the fact that functional items are not

productively used in no way entails that the relevant category is missing since the position

might simply be occupied by a null element (Hyams 1992). However, based on the minimal

overt representation of overt functional elements in the initial stages, and the fact that JC does

not employ other operations defining functional roles (such as subject verb agreement, verb

inversion, case assignment, etc) it is difficult to conclude at this point whether the initial

system in JC lacks a functional projection or not. What we can do however is to provide a

descriptive analysis of the overt production, working our way to a more definitive analysis.

We will now, in more detail, examine the lexical and functional speech categories attested in

the data respectively.

5.3.3. Lexical categories

During the single word stage, according to Radford (1990), children have acquired

phonological and semantic properties but no syntactic properties; hence they are unable to


58

productively combine words into larger structural units. Children’s initial single word

utterances most often include a noun or a bare verb. Although some utterances during the

single word stage (MLU < 1.49) consist of words that correspond to other parts of speech in

the target language, it is nouns which predominate. This is demonstrated in Figure 3 and

corresponding Table 4.

Figure 3: Distribution of Lexical Word Class

MLU (words)

Total lexical Pronouns Nouns Verbs Prepositions Adjectives

< 1.49 1616 157 1053 324 11 71

1.5 – 2.49 12592 2119 5396 3494 551 1032

2.5 – 3.49 30289 7700 9112 8660 2325 2979

3.5 – 4.49 26661 8075 6033 7236 2416 2921

4.5 > 41024 12504 8367 10742 4512 4425

Table 4: Distribution of Lexical Word Class

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

< 1.49 1.5 – 2.49 2.5 – 3.49 3.5 – 4.49 4.5 >

Adjectives

Prepositions

Verbs

Nouns

Pronouns

MLU (in words)

% W

OR

D C

LASS

PR

OD

UC

TIO

N


59

The large production of nouns at the single word stage is not surprising, as it is often reported

that children’s first words are primarily nouns (Gentner 1982, Macnamara 1972, Nelson 1973,

among others). Gentner (1982) provides cross-linguistic evidence of overwhelming agreement

among 6 different languages that the basic pattern of first-word acquisition is concentrated on

nominals. He argues that for semantic reasons, concepts referred to as nouns are particularly

more accessible to infants than verbs or prepositions, because nouns are conceptually more

basic, having transparent mapping to the real world.

In the second stage, MLU 1.5 – 2.49, we note an increase in the percentage production of all

lexical categories with the exception of nouns which demonstrated a major decline from a

high of 65.2% to 42.9%. Verb usage increased from 20% to 27% and thereafter remained

constant throughout all consecutive stages. This stable production of verbal categories from

the start of 2-word combinations suggests that verb usage is already fully acquired and does

not change with development. Adjective8 usage increased gradually levelling off at 10%

during stages 4 and 5 where MLU is over 3.5. Likewise the increase in pronoun usage reached

a steady frequency of 30% during stages 4 and 5. The increase in the production of pronouns

appears to result in the gradual decline of the noun category, reaching a minimum of 20%

during stage 5. Prepositions gradually increased to 11% at stage 5. We will now examine the

three major lexical categories, as revealed by our data, (i.e. nouns, verbs and pronouns) in

more detail.

Closer analysis of nouns in the data (see Table 5) reveals that common nouns are most

attested while, as expected, abstract nouns are virtually unrealized. Proper nouns remained

8 The analysis provided for adjectives also includes predicative adjectives which arguably function like verbs in

JC. A separate analysis was conducted where predicative adjectives were grouped with verbs, but the difference in results was minimal and therefore was excluded from the discussion.


60

fairly constant accounting for 26% – 31% of the nominal utterances, however this declined to

21% during the last stage where MLU is over 4.5.

MLU (words)

Total Nouns

Proper Nouns

Common Nouns

Abstract Nouns

< 1.49 1053 276 773 4

1.5-2.49 5396 1687 3685 24

2.5-3.49 9112 2526 6553 33

3.5-4.49 6033 1569 4436 28

4.5 > 8367 1801 6512 54

Table 5: Distribution of Nouns

From Table 6 it can be seen that there is a very strong pattern in the sequence of production of

verbs. Non-stative verbs are more robustly produced, while the increase in the use of statives

is more gradual. Serial verbs are minimally attested, but will be the subject of a more detailed

discussion in Chapter 9. Locative verbs are also minimally attested.

MLU (words)

Total verbs

Non-statives Statives

Serial verbs

Locative verbs

< 1.49 324 301 10 0 13

1.5-2.49 3494 2960 404 11 119

2.5-3.49 8660 6961 1298 34 358

3.5-4.49 7236 5778 1124 37 292

4.5 > 10742 8757 1580 98 321

Table 6: Distribution of Verbs

With regards to pronouns (Table 7) we note that children start off producing twice as many

first person singular pronouns compared to third person singular pronouns. This confirms

earlier observations that first person singular pronouns are the first to be used in significant

numbers (Chiat, 1986). By stage two however we note a drastic shift in the production of 3rd

person singular pronouns. The data reveals a clear tendency for singular pronouns to emerge

before plurals. Demonstrative pronouns seem to be also minimal in the dataset.


61

MLU (words)

Total Pronoun 1SG 2SG 3SG 1PL 2PL 3PL

Demon-stratives

< 1.49 157 90 9 42 0 0 1 15

1.5-2.49 2119 458 148 1291 2 0 13 207

2.5-3.49 7700 2755 945 3228 8 0 108 656

3.5-4.49 8075 2964 1128 3173 9 0 128 673

4.5 > 12504 5203 2007 4284 32 8 238 732

Table 7: Distribution of Pronouns

The general order of production in terms of frequency of initial lexical production would

therefore be Nouns > Verbs > Pronouns > Adjectives > Preposition.

5.3.4 Functional categories

In the previous sections we saw that the initial utterances of children are mainly lexical, being

attributed to Nouns. These early nominals however appear to lack an overt D-system as only 4

determiners were found during this stage. Likewise, early verbs completely lack an overt I-

system as only 3 inflections were attested. There are absolutely no complementizers produced

in the single word stage. This is demonstrated in Figure 4 and corresponding Table 8.

As shown, the use of complementizers remained relatively unattested throughout the

children’s production. At MLU over 4.5, only 40 complementizers were attested in the data,

representing less than 1% of the functional items produced. This is not very surprising

however as complementizers arise when there are complex sentences, for example relative

clauses and subordinate clauses. If such complex sentences are generally absent from

naturalistic production, then the absence of complementizers in the children’s production

follows.


62

We have seen a gradual decrease in the percentage production of determiners with respect to

the subsequent increase in the use of inflection. At the two latter stages (MLU 3.5 and over)

we note a steady production in determiners and inflection, accounting for 37% and 62%

respectively.

Figure 4: Distribution of Functional Word Class

MLU (words) Total Functional Determiners Complementizers Inflections

< 1.49 7 4 0 3

1.5 – 2.49 229 120 3 106

2.5 – 3.49 2323 1176 3 1144

3.5 – 4.49 2885 1061 9 1815

4.5 > 5895 2188 40 3667

Table 8: Distribution of Functional Word Class

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

< 1.49 1.5 - 2.49 2.5 - 3.49 3.5 - 4.49 4.5 >

Inflection

Complementizer

Determiner

% WORD CLASS PRODUCTI

MLU (in words)


63

The general order of production on the functional categories in terms of frequency would be

Determiners >Inflections > Complementizers.

So far we have been concentrating on the order of emergence and frequency of use of overt

syntactic categories. To complete the account we need to describe the development of

syntactic structures. To do so, we will start by examining the first combinations of word

classes in the following section.

5.4 INITIAL COMBINATIONS

At MLU above 1.5, children start to combine words in systematic patterns, suggesting the

emergence of grammar. The central concern of this section then is when and how children

develop these initial grammars.

MLU 1.5 – 2.49 represents the very first combinations, also described as the two-word stage.

During this stage the main combinations include a verbal and a pronominal element as

displayed in Table 9, and exemplified in (5) to (9). These utterances represent 30% of the

two-word combinations in the declarative utterances attested in the data.

First Combinations Verb + Pronoun

Noun + Noun

Pronoun + Verb

Verb + Noun

Noun + Verb Others

actual utterances 579 333 239 192 172 435

%production 30 17 12 10 9 22

Table 9: Initial Combinations9

9 This analysis is restricted to declarative utterances.


64

5) Kaal ar. (ALA 1;11)

call 3SG

“I am going to call her.”

6) Bai it. (COL 1;08)

buy 3SG

“He bought it.”

7) Toch i. (KEM 2;00)

touch 3SG

“I touched it.”

8) Lik yo. (RJU 1;11)

lick 2SG

“I am going to hit you.”

9) Jraiv dis. (TYA 2;05)

drive this

“I am driving this.”

In addition, Noun Noun combinations, forming the possessive construction, as in 10 – 14, are

also well attested at this 2-words stage.

10) Dora slippaz. (ALA 2;01)

Dora slippers

“Dora’s slippers.”


65

11) Momi kii. (COL 1;11)

Mommy key

“Mommy’s key.”

12) Kevin kyar. (KEM 2;01)

Kevin car

“Kevin’s car.”

13) Manski pensl. (RJU 1;11)

Manski pencil

“Manski’s pencil.”

14) Gorl baal. (TYA 2;06)

Girl ball

“The girl’s ball.”

The next combination attested with some frequency is the use of pronouns followed by verbs

as in 15 – 19.

15) Im baak. (ALA 2;00)

3SG bark

“It barks.”

16) I jrap. (COL 1;09)

3SG drop

“It dropped.”


66

17) Mi wiek-op. (KEM 2;04)

1SG wake.up

“I woke.”

18) Im gaan. (RJU 1;11)

3SG gone

“He has gone.”

19) It brok. (TYA 2;06)

3SG break

“It broke.”

Also evident in the data are Verb Noun strings and Noun Verb strings, the former being more

attested than the latter. Examples are presented in 20 – 24 and 25 – 29 respectively.

20) Want chiiz. (ALA 1;11)

want cheeze

“I want the cheeze.”

21) Wash an. (COL 1;10)

wash hand

“He is going to wash his hand.”

22) Jraiv van. (KEM 2;02)

drive van

“We drive the van.”


67

23) Bai juus. (RJU 2;00)

buy juice

“He is going to buy a juice.”

24) Kola poppi. (TYA 2;06)

colour puppy

“I am going to colour the puppy.”

25) Kiti kom. (ALA 1;10)

kitty come

“The kitty is coming.”

26) Dadi gaan. (COL 1;11)

daddy gone

“Daddy is gone.”

27) Kyar mash-op. (KEM 2;03)

car damage

“The car is damaged.”

28) Dadi bied. (RJU 2;00)

daddy bathe

“Daddy is having a bath.”


68

29) Duor op-op. (TYA 2;05)

door open

“The door is opened.”

A simple analysis of the first combinations suggests that children at this stage are already

aware of lexical word classes and can systematically and consistently combine them in

keeping with the Subject Verb Object (SVO) order attested in the target system. We have seen

a variety of combinations distributed as Subject Verb strings and Verb Object strings; but at

no time have we noted Verb Subject combinations, or other combinations which would

deviate from the target. The initial verb phrase structures demonstrate that children pass

through a stage where they use exclusively bare verbs. Whether these are root infinitives or

not will be the discussed in the next chapter, likewise the subsequent development of

inflection. Additionally, the Verb Object strings appear to be cases of root null subjects and

will be examined in further details in Chapter 7. The Noun Noun structures are evidence of

noun phrases. Thus, the distributional evidence of early word class combinations points to the

onset of categorization and development of grammatical relations. Based on evidence from

longitudinal studies, Radford (1990) points out that once categorization begins in child’s

grammar, it spreads fairly quickly through the child’s lexicon. Turning to utterances of three

or more words, we find evidence that indicates the quick elaboration of phrase structure

formation and the general expansion of grammatical relations.

5.5 MULTIWORD UTTERANCES

In examining combinations of three or more words (represented by the MLU ranging from 2.5

– 3.49 and beyond), we note that not only are there increased combinations of various lexical


69

words but that elements belonging to functional groups are attested with more frequency.

Zooming in on the production of determiners we see a striking increase in their usage, moving

from a mere 120 utterances in the two-word stage (MLU 1.5 – 2.49) to over 1100 utterances

as outlined in Table 10. This development provides evidence of the overt projection of the

basic noun phrases into functional determiner phrases. Closer analysis of the determiners

produced in the data reveals that one essential property of the determiner system at this stage

is the robust use of definite articles in comparison with indefinite articles10

. Additionally

children seem to show awareness of target-consistent distributional properties associated with

determiners; determiners were pre-nominal modifiers of nouns in both subject and object

position, but at no point were there overgeneralizations with a proper noun or with a

pronominal element.

MLU (words) Total Determiners Definite Determiners Indefinite determiners

< 1.49 4 3 1

1.5 – 2.49 120 100 20

2.5 – 3.49 1176 1041 135

3.5 – 4.49 1061 820 241

4.5 > 2188 1753 435

Table 10: Distribution of Determiners

Typical examples of initial definite determiners in the corpus are presented in 30 – 35 and

indefinite determiners in 36 - 41.

30) Di liedi gaan dong. (ALA 2,02)

DEF lady gone down

“The lady went down (the road).”

10 In line with Stewart (2006) indefinite determiners were analyzed as quantifiers.


70

31) Mi a raid i chok. (COL, 2;00)

1SG PROG ride DEF truck

“I am riding the truck.”

32) Ø prie i grong. (KEM 2;05)

Ø spray DEF ground

“He sprayed the field.”

33) Eskyuuz mi lak i duor. (RJU 2;01)

excuse 1SG lock DEF door

“Excuse, I am going to lock the door.”

34) Mi a jraa i waata. (SHU 2;02)

1SG PROG draw DEF water

“I am drawing the water.”

35) Di jres in de. (TYA 2;09)

DEF dress in there

“The dress is in there.”

36) Wan big chok a kom. (ALA 2;03)

INDEF big truck PROG come

“A big truck is coming.”

37) Mi iit a papaa. (COL 2;02)


71

1SG eat INDEF papaya

“I am going to eat a papaya.”

38) Ø waa som juus. (KEM 2;05)

Ø want INDEF juice

“I want some juice.”

39) Wan big kou in de. (RJU 2;03)

INDEF big cow in there

“A big cow is in there.”

40) Mi aa jraa wan cheri chrii. (SHU 2;02)

1SG PROS draw INDEF cherry tree

“I am going to draw a cherry tree.”

41) Mi a luk fi wan mongki. (TYA 2;10)

1SG PROG look for INDEF monkey

“I am looking for a monkey.”

What is suggested here is that one of the defining characteristics of children’s speech during

the initial multiword stage of their development is the acquisition of the grammatical category

of determiners. The children demonstrated the move from a predominantly lexical stage to a

system which constitutes overt functional projections in line with the target system. This

functional development may be also illustrated with the development of the Inflectional

system as presented in Table 11.


72

MLU (words) Total Inflection Aspect Modal Past Tense Infinitive

< 1.49 3 2 0 0 1

1.5 – 2.49 106 100 3 0 3

2.5 – 3.49 1144 1019 96 5 24

3.5 – 4.49 1815 1439 303 23 50

4.5 > 3667 2893 548 43 183

Table 11: Distribution of Inflection

Like for determiners, we note a striking increase in the use of inflections moving from 106

utterances in the two-word stage to over 1100 utterances when initial multi-word utterances

begin. The data suggest that children have moved from a system comprised of a bare VP and

has now projected up to the IP level. The most robust inflectional category present in the

speech of the children is the aspectual system. The development of tense mood and aspect

will be the subject of the following chapter and as such we will not venture into further details

here. What we can see however is the gradual development of the clausal structure from a

mainly lexical representation to a functional system, where inflection is overtly marked in

keeping with the target grammar. The robust production of inflection during this period can be

interpreted as a strong indication that IP is largely available as soon as the earliest multiword

combination begins. The development of functional categories is therefore tied to a relatively

early stage of development.

With regards to the production of overt complementizers, the data does not lend itself to a

comparative analysis of the stage of productive use. We can see however that structures

requiring the overt projection of the CP, such as wh-questions, are also very productive at the

stage where the first multi-word utterances begins, i.e. MLU 2.5 – 3.49 (recall Figure 1 and

Table 2). Wh-questions will be examined in more details in Chapters 7 and 8.


73

5.6 DISCUSSION & CONCLUSIONS

In this chapter clausal structure development and the acquisition of phrases were discussed.

The results so far give initial indications of how the acquisition of syntactic structures of JC

takes place. In examining the developmental path of syntax in our corpus of children, there

was individual variation in the ages and as such a categorization by word-based MLU enabled

a comparative analysis. Based on the data presented, we see that declaratives are the most

robust utterance type and nominals are most frequently produced compared to other word

classes, in particular during the single word stage. It was proposed however that during the

single word stage children do not have syntactic properties (Radford 1990), hence syntactic

development can only be effectively analyzed once productive combination of words into

larger structural units begins.

Looking at the very first two word combinations, certain parallels are evident in the systems

of early syntactic development. First we note the production of mainly bare verbs, unmarked

for temporal specificity and mainly bare nouns, unmarked with respect to nominal specificity.

Thus, parallel to the verbal utterances lacking an overt IP projection, there are nominal

utterances lacking an overt DP projection. This stage however is the earliest characterization

of the language where syntactic developments are initially attested and does not last very long

(but is nonetheless subject to individual variation). We will argue that during this initial stage,

there is the optionality of an overt spell-out of the functional clausal layers, resulting in the

apparent difference between early systems and the target grammar. We will adopt the

Truncation approach of Rizzi (1993/1994) (in line with the modifications of the spell-out

mechanism as detailed in Chapter 7) in order to account for this stage in the grammatical

development of JC. We therefore propose that all the structure is available however children


74

have the option of not spelling out the external layers. This is in contrast to the position of

Radford (1990) and others, who have proposed that early grammars entirely lack functional

projections. Importantly at this initial stage of development, children demonstrate knowledge

of the word order categories and produce the correct SVO order as attested in the target

language.

As the earliest multiword utterances begin, we see the parallel development of determiners

and verbal inflections such as TMA markers. This observation that determiners and

inflections are significantly attested is interpreted as evidence for the existence of a very early

functional stage. Additionally, we note the target-like distribution of the markers at the initial

moment in which they are used. The general quantitative expansion of the various items

within the grammatical categories points to a system of syntactic acquisition that is gradual,

rather than reflecting immediate categorical introduction and/or eradication of the previous

grammatical system. Note that both lexical and functional categories were present from the

single word stage. Additionally, within the functional domain, from the single word state,

there were both overt determiners and inflections in addition to wh-structures requiring the

overt projection of the CP. The gradual increase in words/morphemes in contrast to functional

categories shows that all the functional heads were present from the onset but mature

overtime; thus pointing in favor of the full competence approach of development.

5.6.1 Summary

For the present chapter, based on the step by step descriptive view of the development

patterns of early clauses in JC, we see that initial utterances appear to be a systematic

reduction of adult speech characterized mainly by the omission of overt functional items,


75

attributed to the option of Truncation afforded by children. Initial two-word combinations are

mainly lexical in nature, respecting the distributional rules of the target system. As multi-word

utterances begin and phrase structures become more complicated, the option to truncate

gradually disappears and functional categories are more evident, demonstrating a system

generally in line with the target grammar. The developmental stages and the changes in

acquisition appear to be gradual changes in contrast to immediate categorical eradication of

the previous grammatical system. These facts are in line with a full competence approach to

grammar where there is the existence of an early functional stage.

Chapter 6: The Acquisition of Tense, Modal and Aspect

76

CHAPTER 6

THE ACQUISITION OF TENSE, MODAL AND ASPECT

6.1 INTRODUCTION

This chapter reports on the longitudinal development of inflection as observed in the

spontaneous speech of children acquiring JC, focusing on the production and omission of

Tense, Modal and Aspect (TMA) markers. When used, these markers must intervene between

the subject and the verb, as exemplified in (1) – (3):

1) Mi ben go pan mai chrakta. (KEM 3;00)

1SG PAST go on POSS.1SG tractor

“I went on my tractor.”

2) I shuda fit momi. (ALA 2;07)

3SG MOD fit mommy

“It should’ve fit mommy.”

3) Mi a go11

shuo Jia. (COL 2;09)

1SG ASP show Jia

“I am going to show Jia.”

11 Durrleman-Tame (2008) proposed that progressive a must be used in combination with the prospective go (which may become ao in rapid speech) in order to yield a prospective aspectual meaning. For simplicity purposes, throughout this dissertation, we will treat a+go as a single marker of prospective aspect represented as a_go.


77

In keeping with Durrleman (2000, 2008), inflectional markers in JC are overt manifestations

of clausal functional heads, in line with Cinque’s (1999) functional hierarchy, as presented in

(4):

4) Modepistemic: shuda, wuda, maita, mosa, kuda > T(Past): did/ben T(Future): wi

> Modnecessity: mos > Modobligation: hafi / Modability/permission: kyan > aredi

AspAnterior done1 > still Aspcontinuative > jus Aspretrospective > Aspgeneric/progressive:

a > Aspprospective: go > Aspcompletive: done2 Aspfrequentative: reduplicated verb

How is such a complex functional structure acquired? The transparently analytical

morphological realization of the system makes Creole languages an ideal case to investigate

this question. One conceivable approach is that the TMA zone would emerge incrementally in

a bottom-up manner; we will refer to this view as the ‘incremental structure building’

approach (see Radford 1995, 1996, Clahsen 1996 for variants of this view). On the other

hand, the ‘full competence’ approach (advocated, e.g., in Poeppel & Wexler 1993) posits that

all the structure is available when syntactically significant production starts, but is subject to

grammatical options, not necessarily target-consistent but UG-consistent (such as Truncation)

yielding omissions. We aim to determine the best approach that can account for inflectional

development in child language. The study of JC, being an analytic language, should reveal the

moment at which these inflectional markers first appear and the frequency and productivity of

their use in children’s spontaneous production.

Analyses of the development of TMA are well documented in the literature (Papafragou

(1998), Shirai & Andersen (1995), Bloom et at (1980) on English, Bar-Shalom (2002) on

Russian, Lorusso (2007) on Italian, Panitsa (2010) on Modern Greek, Lee (2009) on Korean,

etc.). However, there are no studies to date on the acquisition of TMA in JC. It is our aim to


78

investigate the development of the TMA system in JC, thereby providing evidence for the

moment at which the first markers appear, the frequency in which the varying forms are used

and their order of acquisition.

The chapter is organized as follows. First we provide an overview of TMA in adult JC. We

then detail the development of the modal zone and examine the target-inconsistent omissions

in Section 6.3. Subsequently, we show the difficulty in establishing the point of acquisition of

the past tense feature, based on the optionality in overtly marking the verb for past tense

reading and the sporadic use of the overt past tense markers. In Section 6.5 we examine

Aspects. The discussion primarily surrounds the progressive and prospective aspectual

markers as the other markers are minimally attested. Along similar lines as the analysis

presented for modals, we examine the target-inconsistent omissions, revealing what appears to

be a root infinitival stage in JC which becomes the subject of Section 6.6. We present in

Section 6.7 a general overview of the entire TMA zone examining the interactions between

tense, modal and aspect in the child production then making comparisons with their

distribution in the input. The paper ends with a discussion of the main findings thereby

proposing an approach which may best account for inflectional development in JC.

6.2 AN OVERVIEW OF TENSE, MODAL & ASPECT IN JC

Jamaican Creole does not have bound verbal inflections but there are various free morphemes

that accompany the verb to express Tense, Modality and Aspect. These markers all occur

preverbally. The preverbal marker for past in JC is did or ben (with variants wen and en).

According to Patrick (2007), these markers occur more rarely than the classic creole pattern

predicts. An unmarked non-stative verb tends to yield a default past reference, while stative


79

verbs tend to yield a non-past interpretation, however other factors may impact the temporal

reading, including context, a time-adverbial or the nature of the direct object (Durrleman-

Tame 2008).

An unmarked verb alone cannot express progressive aspect, though it can express habitual

aspect. The preverbal progressive aspect marker a, da or de, must be used to indicate

progressive action. The past markers may combine with the progressive markers to give

bena/bende, dida, and wena with a meaning corresponding to English past progressive.

Completive aspect is signaled by don in JC. It can appear either before or after the verb

phrase, but is restricted to non-statives. Where don occurs pre-verbally without the stativity

restriction, it gives an anterior reading (Durrleman-Tame 2008). The marker jos is used to

designate retrospective aspect, along the lines of Standard English ‘just’ (see also Cinque

1999). Futurity may be expressed by a_go and wi in JC. A_go (with variants aa, goo, gaa(n),

gwain) is employed to express prospective aspect or proximal future. Wi has been analyzed as

future tense (Durrleman-Tame, 2008) and as a modal (Bailey 1966).

JC has an extensive group of modals which occurs in various combinations with each other.

Bailey (1966:45, 141) divides them into two groups by order of occurrence:

Mod 1: kuda ‘could’, shuda ‘should’, wuda ‘would’, maita ‘may, might’ and wi ‘will’

Mod 2: hafi ‘have to’, mos(a), ‘must’ kyan ‘can’ and fi ‘ought’

We will now examine in detail the phenomenon of modality in JC.


80

6.3 MODALITY

Modals in JC are divided into two classes, epistemic modals and root modals. Epistemic

modals are concerned with the speakers’ deductions or opinions and are situated high in the

clausal hierarchy, while root modals are subject oriented and are situated in the lower domain

(see Cinque 1999). Following Durrleman-Tame (2008) root modals are further divided into 3

classes: ability/permission modals, obligation modals and necessity modals.

As briefly mentioned in the previous section, there appears to be some differences in the

literature regarding the dual classification of wi as future tense (Durrleman-Tame 2008) and

modal (Bailey 1966). According to Van de Vate (2011; 194) ‘… the debate on whether future

time reference is a Tense category or Modal category has a long tradition, and the outcome

has not yet been settled.’ In the literature, a modal analysis for morphemes which indicate a

future tense interpretation is not uncommon (Huddleston 1995, Copley 2002, Werner 2003,

Matthewson 2006, Van de Vate 2011, among others). We will therefore analyse wi within the

functional domain of modality in line with Bailey (1966).

The inventory of modals for the current work is therefore as follows:

5) Epistemic: shuda~shudn, wuda~wudn, kuda~kudn, mosa/mosi, maita, wi12

6) Root Ability/permission: kyan~kyahn, Necessity: mos, Obligation: afi

12

In JC, as a marker of modality, wi ranges from root modal (expressing volition/intention as in (1)) to epistemic modal (expressing prediction as in (2)). However for the purposes of this analysis wi will be analyzed as a single category ‘Modal wi’ being distinct from epistemic modals. 1) Mi wi fiks it. (SHU 2,11) 1SG MOD fix 3SG ‘I will (am willing to/intend to) fix it.’ 2) I wi muuv ahn slak op. (KEM 3,01) 3SG MOD move and slack up ‘It will move and become slack.’


81

Note that the alternation as indicated by ~ refers to the negated variant of the respective

marker.

6.3.1 Acquisition of Modality

A number of studies focusing on modality (Brown 1973, Kuczaj and Maratsos 1975, among

others, reviewed in Papafragou 1998) all concluded that English modals emerge between 1;10

and 2;06. Wells (1979) found that epistemic modality is acquired later than root modality. He

notes that by 2;06, more than 50% of the children in his sample used ‘can’ to convey both

ability and permission; by the same time children used ‘will’ to communicate intention. Wells

(1985) reports that between 2;09 and 3;00, children use ‘must’, ‘have to’ and ‘should’ to

communicate obligation or necessity, but, unlike ‘can’ and ‘will’, these uses did not reach

steady frequencies until around 3;03. There seems to be a general conclusion (Pea, Mawby

and MacCain (1982), Kuczaj & Maratsos (1975) Perkins (1983) Stephany (1979/1986, cited

from Papafragou, 1998)) pointing to the late development of epistemic modality. A likely

explanation put forward is the role of input: most modal expressions produced by parents to

children are those related to permission, ability, obligation, and other related notions.

We will now examine this phenomenon in JC.

6.3.2 Acquisition of Modality in JC

The data provides evidence in support of the claim that modality is generally produced rather

late - after 2 years of age (in keeping with Radford (1990) and others for English). In this

section we examine the order of acquisition of the modal markers and how they fit in the

functional projections of the Inflectional Phrase.


82

Based on the data presented in Tables (1) – (6), the ability modal kyan and its negative variant

kyahn, is the first modal to be produced by all the children in our corpus, with the exception

of ALA, who seems to have had a single sporadic production of modal wi at 2;2. These ability

modals all emerged after the 2nd

birthday, between 2;1 and 2;3, with the exception of RJU

whose first production was at 1;11 and TYA whose first production was much later at 2;10.

ALA

AGE(Y;M,D) ABL OBL NEC EPIS WI

1;9,25 0 0 0 0 0

1;10,4 0 0 0 0 0

1;10,25 0 0 0 0 0

1;11,5 0 0 0 0 0

1;11,16 0 0 0 0 0

2;0,9 0 0 0 0 0

2;0,20 0 0 0 0 0

2;1,5 0 0 0 0 0

2;1,20 0 0 0 0 0

2;2,6 0 0 0 0 0

2;2,22 0 0 0 0 1

2;3,8 3 0 0 0 0

2;3,24 1 0 0 0 0

2;4,9 1 0 0 0 0

2;4,24 0 1 0 0 0

2;5,7 9 0 0 0 0

2;5,23 7 0 0 1 1

2;6,12 2 1 0 0 0

2;6,22 1 1 1 0 0

2;7,5 9 0 0 1 0

2;7,18 7 3 3 3 0

2;8,2 11 5 0 0 0

2;8,16 22 0 0 2 1

2;9,0 7 5 0 2 1

2;9,14 30 0 1 0 1

2;9,28 19 5 0 1 1

2;10,25 18 4 0 1 1

2;11,1 22 0 0 1 0

2;11,18 10 4 0 6 0

3;0,1 5 3 0 5 2

3;0,15 17 11 0 6 0

TOTAL 201 43 5 29 9

Table 1: COL Modal Production Table 2: ALA Modal Production

COL


1;8,17 0 0 0 0 0

1;8,27 0 0 0 0 0

1;9,17 0 0 0 0 0

1;9,28 0 0 0 0 0

1;10,8 0 0 0 0 0

1;11,1 0 0 0 0 0

1;11,12 0 0 0 0 0

1;11,28 0 0 0 0 0

2;0,12 0 0 0 0 0

2;0,28 0 0 0 0 0

2;1,14 0 0 0 0 0

2;2,0 0 0 0 0 0

2;2,16 0 0 0 0 0

2;3,1 0 0 0 0 0

2;3,16 1 0 0 0 0

2;3,30 1 0 0 0 1

2;4,15 0 0 0 0 1

2;5,0 0 0 0 0 0

2;5,14 8 1 0 0 1

2;5,27 1 0 0 0 0

2;6,10 0 0 0 0 3

2;6,25 1 0 0 0 0

2;7,8 5 0 0 0 0

2;7,22 1 0 0 0 1

2;8,6 2 0 1 0 0

2;8,20 31 0 0 0 0

2;9,11 2 0 0 0 0

2;9,24 7 0 0 0 0

2;10,10 13 0 0 0 0

2;10,21 18 0 0 0 0

2;11,7 30 0 0 0 0

TOTAL 121 1 1 0 7


83

Table 3: RJU Modal Production Table 4: TYA Modal Production

RJU


1;10,4 0 0 0 0 0

1;10,14 0 0 0 0 0

1;11,4 0 0 0 0 0

1;11,15 0 0 0 0 0

1;11,26 1 0 0 0 0

2;0,19 0 0 0 0 0

2;0,30 1 0 0 0 0

2;1,15 3 0 0 0 0

2;2,0 2 0 0 0 0

2;2,16 0 0 0 0 0

2;3,1 1 0 0 0 2

2;3,18 5 0 0 0 1

2;4,4 0 0 0 1 0

2;4,19 1 0 0 0 0

2;5,3 1 0 0 0 0

2;5,17 1 0 0 0 0

2;6,2 5 0 0 0 0

2;6,18 5 0 0 0 1

2;7,5 12 0 0 0 1

2;7,15 10 0 0 0 1

2;7,28 6 0 0 0 3

2;8,12 16 0 0 0 1

2;8,26 10 0 1 0 0

2;9,10 7 0 0 0 0

2;9,24 4 0 0 0 0

2;10,7 2 0 0 0 0

2;10,29 5 0 0 0 0

2;11,11 1 0 0 0 0

2;11,25 7 1 0 0 2

3;0,11 6 0 0 0 0

3;0,25 12 1 2 0 2

TOTAL 124 2 3 1 14

TYA


1;11,25 0 0 0 0 0

2;0,4 0 0 0 0 0

2;0,25 0 0 0 0 0

2;1,5 0 0 0 0 0

2;1,16 0 0 0 0 0

2;2,9 0 0 0 0 0

2;2,20 0 0 0 0 0

2;3,5 0 0 0 0 0

2;3,20 0 0 0 0 0

2;4,6 0 0 0 0 0

2;4,22 0 0 0 0 0

2;5,8 0 0 0 0 0

2;5,24 0 0 0 0 0

2;6,9 0 0 0 0 0

2;6,24 0 0 0 0 0

2;7,7 0 0 0 0 0

2;7,23 0 0 0 0 0

2;8,8 0 0 0 0 0

2;8,22 0 0 0 0 0

2;9,5 0 0 0 0 0

2;9,18 0 0 0 0 0

2;10,2 1 0 0 0 0

2;10,16 0 0 0 0 0

2;11,0 4 0 0 0 0

2;11,14 0 0 2 0 0

2;11,28 7 0 0 0 0

3;0,19 0 0 0 0 0

3;1,1 10 0 0 0 1

3;1,15 11 1 0 0 0

3;2,1 8 0 0 0 0

3;2,15 16 1 0 0 0

TOTAL 57 2 2 0 1


84

SHU


2;1,23 0 0 0 0 0

2;2,2 0 0 0 0 0

2;2,11 0 0 0 0 0

2;3,3 0 0 0 0 0

2;3,14 1 0 0 0 0

2;4,7 0 0 0 0 0

2;4,18 0 0 0 0 0

2;5,3 0 0 0 0 0

2;5,18 8 0 0 0 0

2;6,4 4 1 0 0 0

2;6,20 1 0 0 0 1

2;7,16 5 0 0 0 1

2;7,22 4 0 0 0 0

2;8,7 7 0 0 0 0

2;8,22 3 0 0 0 0

2;9,5 2 0 0 0 0

2;9,21 4 0 0 0 0

2;10,6 1 0 0 0 0

2;10,20 1 0 0 0 0

2;11,3 1 0 0 0 1

2;11,16 4 0 1 0 0

3;0,0 7 0 1 1 0

3;0,19 6 0 1 0 0

3;0,28 4 2 1 0 0

3;1,12 1 1 1 0 0

3;1,26 18 2 3 0 0

3;2,17 7 0 0 0 0

3;2,30 19 2 2 0 0

3;3,16 1 2 1 0 0

3;3,27 11 2 0 0 0

3;4,13 7 4 1 0 0

TOTAL 127 16 12 1 3

Table 5: KEM Modal Production Table 6: SHU Modal Production

KEM


2;0,21 0 0 0 0 0

2;1,0 0 0 0 0 0

2;1,21 1 0 0 0 0

2;2,1 0 0 0 0 0

2;2,12 0 0 0 0 0

2;3,5 0 0 0 0 0

2;3,16 0 0 0 0 0

2;4,1 0 0 0 0 0

2;4,16 0 0 0 0 0

2;5,2 0 0 0 0 0

2;5,18 0 0 0 0 0

2;6,4 0 0 0 0 0

2;6,20 2 0 0 0 0

2;7,5 21 0 0 0 0

2;7,20 5 0 0 0 0

2;8,3 0 0 0 0 1

2;8,19 0 0 0 0 1

2;9,4 0 0 0 1 0

2;9,18 4 1 0 0 2

2;10,1 4 0 0 0 0

2;10,14 9 0 0 1 0

2;10,29 7 0 0 0 2

2;11,12 4 0 0 0 0

2;11,26 17 0 0 0 1

3;0,10 2 0 1 0 1

3;0,24 3 1 0 0 3

3;1,15 11 2 0 0 3

3;1,28 22 4 0 0 5

3;2,11 10 2 0 0 4

3;2,25 5 1 0 0 4

3;3,11 0 0 0 0 0

TOTAL 127 11 1 2 27


85

Table 713

details the total production of all the modals in the corpus. As can be seen, the

ability/permission modal is the most frequently produced, accounting for 80% (757 of 950) of

the utterances expressing modality. The root obligation modal accounts for 8% (75 utterances)

and modal wi accounts for 6% (61 utterances). The necessity and epistemic modals each

accounts for roughly 3% of the data set. From a holistic analysis, as presented in Table (7) it

would appear that the ability modal was first produced at 2;0, modal wi comes on stream at

2;2.5 and the epistemic modal follows at 2;4. However, looking back at Tables (1) – (6) we

note there are huge individual variations with the order and frequency of production of the

modals. With regards to the production of epistemic modals, the data does not lend itself to a

cross-sectional examination of the phenomenon as 87% (29 of 33 utterances) was produced

by one informant, ALA. It was however first produced by RJU (his sole expression of

epistemic modality) at 2;4. ALA’s first epistemic utterance was at 2;5.5 months.

13

In order to present a cumulative analysis, approximate age was used for the participants. For e.g. 1;9.5 means 1 year 9 and half months.


86

TOTAL

AGE(Y;M) ABL OBL NEC EPIS WI

1;8.0 0 0 0 0 0

1;8.5 0 0 0 0 0

1;9.0 0 0 0 0 0

1;9.5 0 0 0 0 0

1;10.0 0 0 0 0 0

1;10.5 0 0 0 0 0

1;11.0 0 0 0 0 0

1;11.5 0 0 0 0 0

2;0.0 1 0 0 0 0

2;0.5 0 0 0 0 0

2;1.0 1 0 0 0 0

2;1.5 4 0 0 0 0

2;2.0 2 0 0 0 0

2;2.5 0 0 0 0 1

2;3.0 5 0 0 0 2

2;3.5 8 0 0 0 2

2;4.0 1 0 0 1 1

2;4.5 1 1 0 0 0

2;5.0 18 1 0 0 1

2;5.5 17 0 0 1 1

2;6.0 11 2 0 0 3

2;6.5 10 1 1 0 2

2;7.0 52 0 0 1 2

2;7.5 27 3 3 3 2

2;8.0 26 5 1 0 4

2;8.5 72 0 0 2 3

2;9.0 21 5 1 3 1

2;9.5 52 1 1 0 3

2;10.0 42 5 0 1 1

2;10.5 48 4 0 2 1

2;11.0 69 0 0 1 3

2;11.5 19 4 3 6 0

3;0.0 43 4 1 6 5

3;0.5 31 11 2 6 1

3;1.0 29 4 3 0 6

3;1.5 23 4 1 0 3

3;2.0 48 6 3 0 5

3;2.5 33 3 0 0 4

3;3.0 24 3 2 0 4

3;3.5 1 2 1 0 0

3;4.0 11 2 0 0 0

3;4.5 7 4 1 0 0

TOTAL 757 75 24 33 61

Table 7: Total production of Modality

We present examples of each utterance type produced.


87

Ability/permission modal

7) Ø kyan brok dis pliiz? (ALA 2;03)

Ø MODabl break DEM please

“Can you break this please?”

8) I kyahn spin. (COL 2;05)

3SG MODabl spin

“It cannot spin.”

9) Ø kyahn go fors, wiet pahn mi! (KEM 2;07)

Ø MODabl go first wait on 1SG

“You cannot go first, wait on me!”

Obligation modal

10) Mi afi muuv di baisikl rait yaso. (RJU 3;01)

1SG MODobl move DET bicycle right LOC

“I have to move the bicycle right here.”

11) Yaa-fi go bai wan biga wan. (SHU 2;06)

2SG~MODobl go buy Q:indef bigger one

“You have to buy a bigger one.”

12) Chrii fi ina griin. (TYA 3;01)

tree MODobl into green

“Trees have to be (coloured) in green.”


88

Necessity modal

13) Wen shi a kum dong yo mos kom ya

REL 3SG PROG come down 2SG MODnec come LOC

kom luk fi ar yu ier? (SHU 3;04)

come look for 3SG 2SG hear

“When she is coming down you must come here to visit her, do you hear?”

14) A mos skid i oot. (COL 2;08)

1SG MODnec skid 3SG out

“I must skid it out.”

15) Im mos it i. (ALA 2;06)

3SG MODnec eat 3SG

“He must eat it.”

Modal wi

16) Di naïf wi kot yo. (KEM 3;00)

DET knife MODwi cut 2SG

“The knife will cut you.”

17) Mi wi fiks it. (SHU 2;11)

1SG MODwi fix 3SG

“I will fix it.”

18) I wi jrap ina di tangk. (RJU 2;03)

3SG MODwi drop into DET tank


89

“He will fall in the tank.”

Epistemic Modal

19) I shuda fit momi. (ALA 2;07)

3SG MODepis fit mommy

“It should’ve fit mommy.”

20) A woda afi get som jakit. (ALA 2;10)

1SG MODepis MODnec get Q:indef jacket

“I would have to get some jackets.”

21) Mosi im a_go jraiv di kyar. (ALA 3;00)

MODepis 2SG PROS drive DET car

“Maybe he is going to drive the car.”

In keeping with Boland (2006) frequency in use of TMA markers does not show how

productive they are, and as such qualitative analyses are needed. One such analysis is their

variation with predicates. There is no general agreement however regarding the number of

different predicates a marker must be used with to be considered productive. A criterion of

two different predicates is used in some research (Pizzuto & Caselli, 1994); however Boland

(2006) posits that to be a very low standard and thereby sets his criterion to five different

predicates. Being that the criteria are quite arbitrary, we will present 2 separate analyses, first

assuming 2 predicates and then 5 predicates as a standard threshold. Tables 8 and 9 show the

age at which the markers are used productively with 2 and 5 different predicates respectively.


90

MODAL COL ALA RJU TYA KEM SHU

ABI/PER 2;3,30 2;3,8 2;0,30 2;11,0 2;7,5 2;5,18

OBL - 2;6,12 3;0,25 - 3;1,15 3;0,28

NEC - 2;7,18 3;0,25 - - 3;0,0

EPIS - 2;6,5 - - - -

WI 2;4,15 2;5,23 2;6,18 - 2;8,19 2;11,3

Table 8: Use of Modals with 2 different predicates

MODAL COL ALA RJU TYA KEM SHU

ABI/PER 2;5,14 2;5,7 2;2,0 2;11,28 2;7,20 2;6,4

OBL - 2;7,18 - - 3;1,28 3;1,26

NEC - 2;9,14 - - - 3;1,12

EPIS - 2;9,0 - - - -

WI 2;6,10 2;9,28 2;7,28 - 3;0,10 -

Table 9: Use of Modals with 5 different predicates

The data reveals that the ability/permission modal is clearly the first modal marker to be used

productively by all the informants, whether we assume the 2 or 5 predicates criterion. We see

however when the threshold is set to 2 predicates, the second marker to be productively used

is the modal wi. The order of the productive use of the other markers seems to be individually

determined. ALA is the only child to use all the markers productively, including the epistemic

modal which is yet to be used productively by any of the other children.

6.3.3 Order of Acquisition of Modal Markers

Based on the individual variation and the sporadic occurrences of the modal elements in the

corpus, the order in which the forms are acquired cannot be firmly established. It is clear

however that the ability/permission modal is the first to be produced and used productively by

children acquiring JC followed by modal wi. Closer examination of the corpus reveals that


91

while there are 3922 child directed utterances using the ability/permission modals, the other

categories are minimally attested, as demonstrated in Figure 1. There are only 714 cases of the

obligation modal, 403 of the necessity modal, 340 of the epistemic modal and 679 modal wi.

This minimal use in the input of other modals to those expressing ability/permission may help

to account for the sporadic production of these other modals in the acquisition data.

Figure 1: Distribution of Modality in Child Production and Input

The data in Figure 1 shows a striking correspondence between the distribution of the modals

in the input and in the children’s productions. With the exception of the ability/permission

modal, the relative difference between the input data and the children’s utterances ranges

merely between 3% and 5%. The distribution of the ability/permission modal in the children’s

utterances is higher than that attested in the input, while all other modality markers are

smaller. This suggests that the ability/permission modal is indeed easier to acquire than the

other modals, and hence is the first to be acquired and productively used. We return to a

discussion on the role of input in section 6.7.2.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

ABL OBL NEC EPIS WI

CHILD

INPUT


92

6.3.4 Omission of Modals

There are some omissions of modals in the corpus: 88 ability, 72 modal wi, 1 obligation, and

3 necessity modals were omitted. These omissions are detected based on the context of the

utterance as exemplified in (22) - (23) below:

22) CHI: we yo kyari fi mi?

“What did you carry for me?”

INV: we mi kyari fi yo, mi kyari wahn buk an wan pensl.

“What did I carry for you, I carried a book and a pencil.”

CHI: mm?

“mm?”

INV: wan buk ah wan pensl, fi yo rait ina, ahn fi yo kyari go skuul.

“A book and a pencil, for you to write in, and for you to carry to school.”

CHI: mi Ø sii it? (COL 2;08)

1SG ØMODabl/per see 3SG

“Can I see it?”

23) INV: si ans de so tu, yo si dem?

“See ants are there also, do you see them?”

CHI: dem Ø bait yo ino. (RJU 3;00)

3PL ØMODwi bite 2SG you~know

“They will bite you, you know.”

CHI: no toch dem!

“Don’t touch then!”


93

INV: mmhm, mi naa toch dem.

“No, I’m not touching them.”

We should point out however, that some of these cases included in the omissions of the

ability modal may be somewhat controversial, as target-consistent interpretations could be

arrived at even with the omissions of the modals. This is exemplified in (24):

24) Context: Child looking at a background display photo on a mobile phone.

CHI: uu dat, a Kyahri ?

“Who is that, is it Kyahri?”

INV: a Kyahri.

“It is Kyahri.”

CHI: (Ø Ø) mek mi si im? VS mek mi si im!

(Ø ØMODabl/per) make 1SG see 3SG make 1SG see 3SG

“(Can you) let me see him?” “Let me see him!”

INV: yo wa si im, pres it an luk if yo si im.

“You want to see him, press it and look if you see him.”

We argue however that based on the intonation associated with these utterances they are

better interpreted as interrogatives and not as imperatives. As interrogatives, they necessarily

require the subject and the modal. As such, we analyse these structures as cases of modal (and

subject) omission.


94

6.3.5 Intermediary Conclusion

In early JC, various types of utterances expressing modality are produced. These include

ability/permission modals, obligation modals, necessity modals, epistemic modals and

prediction/intention modal wi. These modality markers are generally acquired after 2 years of

age. The earliest type of modality to be produced and used productively is the

ability/permission modal. This is expected as it is the most widely used modal – for requests,

assertions and interrogations. As for the other modals, based on individual variation and their

sporadic attestation, we are unable to put forward conclusive evidence as to the order of their

acquisition. The following sequence however is their first attestation in the corpus: ModAbility>

Modwi> ModEpistemic & ModObligation> ModNecessity. We can safely conclude that until the third

year of life, children acquiring JC are still yet to master the acquisition of various modal

categories.

We will now focus our discussion on tense.

6.4 TENSE

6.4.1 Acquisition of Tense

It has been consistently observed that in L1 acquisition of English, children initially use past

marking on accomplishment and achievement verbs more frequently than on activity and

stative verbs. There is however considerable disagreement concerning the phenomenon, both

at the level of description and at the level of explanation (see Shirai & Andersen 1995 for a

review). Some studies report that past marking is only given to actions with clear end results


95

(e.g. Antinucci & Miller 1976, Bronckart and Sinclair, 1973 cited in Shirai & Andersen

1995). These authors suggested that the non-normative use of the tense marking by children

was attributed to a cognitive deficit, where children did not have the concept of tense.

Bickerton (1981) on the other hand interpreted these results as dependent on the punctual-

nonpunctual distinction which he claimed forms an integral part of the core grammar of all

children. He proposed, in line with his Language Bioprogram Hypothesis, that nonpunctual

events are those that have measurable duration or are repeated, and are generally marked

while punctual events are single, completed events which are unmarked.

Bloom et al. (1980) claimed that aspect is acquired before tense. Weist et al. (1984) on the

acquisition of Polish, claimed that children mark both tense and aspect at early stages, thus

providing counter examples to the ‘aspect before tense’ hypothesis. The variance in methods

and procedures employed in different studies can however create differences in interpretation,

and thereby generates problems for comparative analysis. In the following section, we will

examine the production of tense markers in JC.

6.4.2 Acquisition of Tense in JC

The data reveals 5765 occurrences of unmarked verbs with a past time interpretation as

exemplified in (25) – (27), and 4404 unmarked verbs with a present interpretation as

exemplified in (28) – (30).

25) Moesha du dem. (TYA 2;06)

Moesha do 3PL

“Moesha did them.”


96

26) Mi kuk ih kou. (RJU 2;03)

1SG cook DET cow

“I cooked the cow.”

27) Dadi bai i fi mi. (SHU 2;02)

Daddy buy it for me

“Daddy bought it for me.”

28) Mi waa tek yo picha. (ALA 2;05)

1SG want take 2SG picture

“I want to take your picture.”

29) I luk priti. (COL 2;05)

3SG look pretty

“It looks pretty.”

30) Iih av iih baik. (KEM 2;09)

3SG have 3SG bike

“He has his bike.”

Based on the option of using the unmarked verb to express a past or present reading, it is

impossible to determine at what exact point children acquire the concept of tense. Children

rarely select the option of using the overt past tense markers, as of the 5836 utterances with a

past time interpretation only 71 overtly marked past tense markers were found in the entire

corpus. This is detailed in Table 10. The minimal use of the overt markers is expected, as


97

these markers also occur rarely in the speech of adults. Only 1765 overt past markers were

attested in the input data of the corpus.

AGE (Y;M) COL ALA RJU TYA KEM SHU

2;1.0 1 0 0 0 0 0

2;1.5 0 0 0 0 0 0

2;2.0 0 0 0 0 0 0

2;2.5 1 0 0 0 0 0

2;3.0 0 0 0 0 0 0

2;3.5 0 0 0 0 0 0

2;4.0 0 0 0 0 0 0

2;4.5 1 0 1 0 0 0

2;5.0 0 0 0 0 0 0

2;5.5 0 0 0 0 0 0

2;6.0 0 1 0 0 0 0

2;6.5 0 0 0 0 0 1

2;7.0 0 7 0 0 0 0

2;7.5 0 3 2 0 0 0

2;8.0 0 0 0 0 0 2

2;8.5 1 2 4 0 0 0

2;9.0 0 2 0 0 0 1

2;9.5 0 0 0 0 0 0

2;10.0 0 0 0 0 0 0

2;10.5 0 0 0 0 0 1

2;11.0 0 1 0 0 1 1

2;11.5 0 2 0 0 0 0

3;0.0 0 2 0 0 1 0

3;0.5 0 0 0 0 0 2

3;1.0 0 0 0 0 10 0

3;1.5 0 0 0 0 3 2

3;2.0 0 0 0 0 1 4

3;2.5 0 0 0 0 1 0

3;3.0 0 0 0 0 0 3

3;3.5 0 0 0 0 3 0

3;4.0 0 0 0 0 0 1

3;4.5 0 0 0 0 0 2

TOTAL 4 20 7 0 20 20

Table 10: Production of overt tense markers

The data reveals much individual variation with regards to the use of the overt markers: while

there are 20 utterances by ALA, SHU, and KEM, it is never attested in the production of TYA


98

and virtually unattested in COL’s and RJU’s production. There are also variations in the age

of the informants regarding their first production of the overt past tense forms.

Closer examination of the data reveals that of the 71 utterances, 38 were actually expressing

past-progressive, as in (31) as compared to only 33 which yields a simple past interpretation

as in (32):

31) Mi ben a kaal yo. (KEM 3;00)

1SG PAST PROG call 2SG

“I was calling you.”

32) A dis did kom an mi an. (ALA 3;00)

FOC DEM PAST come on 1SG hand

“It is this that came on my hand.”

As done for the modals, presented in Tables 11 and 12 is the age where the past tense marker

is used productively with 2 and 5 predicates respectively, both with a past progressive

interpretation and in its bare form. It is shown that ALA is the first child to use the past-

progressive marker productively in keeping with both the 2 and 5 predicate criterion, while it

never gained productive use in COL’s and TYA’s productions. COL was the first informant to

use the bare past productively based on the 2 predicate criterion, but except for KEM, its

productive use is unattested with the 5 predicate treshold. We will return to a discussion of the

combination of the overt past marker and the progressive aspect marker in Section 6.7.1.

TENSE COL ALA RJU TYA KEM SHU

PAST-PROG - 2;7,5 2;8,12 - 3;0,24 2;11,3

PAST 2;5,0 2;11,1 2;8,12 - 3;0,24 2;9,5

Table 11: Use of overt past tense with 2 different predicated


99

TENSE COL ALA RJU TYA KEM SHU

PAST-PROG - 2;7,18 - - 3;3,11 3;2,30

PAST - - - - 3;0,24 -

Table 12: Use of overt past tense with 5 different predicated

In order to verify the claim that children initially use past marking on accomplishment and

achievement verbs more frequently than on activity and stative verbs, we examined all the

verbs with which the children produced the overt past marker, which demonstrate only a past

reading. In keeping with the literature, we see that the production of the past marker with

verbs of accomplishment and achievement (such as fiks ‘fix’, fain ‘find’, slaid ‘slide’,etc.) is

more frequently attested than with verbs of activity and states.

The huge individual variation and the limited production of the overt past tense markers, does

not lend itself to an indepth comparative analysis of the acquisition of the phenomenon in the

present corpus. Nonetheless, despite the scarcity of the data, the data becomes relevant for the

‘aspect before tense hypothesis’, which we discuss in the following section. We will now

examine aspects, the zone following tense in the functional hierarchy.

6.5 ASPECT

In JC aspectual markers form the group of inflectional particles located closest to the VP.

They do not occur with stative verbs (Lamiroy, 1987: 284). Data presented in Durrleman-

Tame (2008) uphold that the aspectual zone in JC is highly articulate, involving various

projections that host functional material. The structure arrived at for the aspectual zone is as

follows:

33) Asp [retrospective] > Asp [progressive] > Asp [prospective] > Asp [completive]


100

6.5.1 Acquisition of Aspect

There are varying accounts in studies of children’s acquisition of aspects in terms of the age

of acquisition of the aspectual zone, the order of the acquisition of different aspectual

expressions, the frequency of production and the interaction between aspect and tense (See

Andersen and Shirai (1996) for discussion). Shirai and Andersen (1995) put forward that

children acquiring English first use progressive marking mostly with activity verbs, then

extending it to accomplishment and achievement verbs. Additionally, children do not

incorrectly overextend progressive markings to stative verbs. This, Bickerton (1981) claimed

is because children are born with what he calls the ‘state-process distinction’. He argues that

children have innate knowledge of the difference between process and state: process verbs, in

contrast to stative verbs, are marked with the help of a non-punctual marker. He argues that if

children did not possess this knowledge, they would generalize the progressive marker just as

they do past and plurals in English. Moreover, Bickerton proposed that creole languages

exhibit striking evidence of linguistic universals in the area of tense and aspect. He states that

there are at least four basic binary semantic distinctions which form an integral part of the

core grammar which constitutes the totality of preexperiential linguistic knowledge. Three of

these distinctions are related to the TMA system and appear to be shared by almost all creoles.

We will now present a detailed examination of aspect in JC.

6.5.2 Acquisition of Aspect in JC

Presented below are examples of target-consistent use of aspectual marker in the corpus:


101

Completive:

34) Di rien don faal. (ALA 3;00)

DET rain COMP fall

“The rain stopped falling (completely).”

35) Manski don iit fi ar aredi. (RJU 2;11)

Manski COMP eat POSS 3SG already

“Manski already finished eating hers (completely).”

36) A no don kuul i. (SHU 3;00)

1SG NEG COMP cool 3SG

“I havent finished cooling it (completely).”

Progressive:

37) Mi a iit i. (COL 2;00)

1SG PROG eat 3SG

“I am eating it.”

38) Wi a kom. (KEM 2;04)

1PL PROG come

“We are coming.”

39) Mi a wash mi fut gud. (TYA 2;08)

1SG PROG wash 1SG foot good

“I am washing my foot properly.”


102

Prospective:

40) Dadi a_go bied. (COL 2;00)

Daddy PROS bathe

“Daddy is going to have a bath.”

41) Mi a_go sliip now. (KEM 2;06)

1SG PROS sleep now

“I am going to sleep now.”

42) Mi aa rait it. (TYA 2;05)

1SG PROS write 3SG

“I am going to write it.”

Retrospective:

43) Shi jos kaal mi. (SHU 2;11)

3SG RETRO call 1SG

“She just called me.”

44) Im jos bait mi. (ALA 2;06)

3SG RETRO bite 1SG

“He just bit me.”

45) I jos mash-op. (RJU 2;07)

3SG RETRO mash up

“It was just damaged.”


103

Detailed in Tables 13 – 18 is the individual production of overt aspect markers unfolding the

development of the aspectual zone in the corpus. The data reveals that the progressive aspect

marker a is the first to be produced, followed closely by the prospective a_go. We see a

steady increase of overt markings of both the progressive and prospective aspectual markers

as the age of the informants increases. The production of the completive and retrospective

markers appears to be more sporadic and individual variation is significant.

ALA

AGE(Y;M,D) COMP PROG PROS RETRO

1;9,25 0 0 0 0

1;10,4 0 1 0 0

1;10,25 0 1 0 0

1;11,5 0 6 0 0

1;11,16 0 1 0 0

2;0,9 0 4 0 0

2;0,20 0 3 0 0

2;1,5 0 3 0 0

2;1,20 0 5 0 0

2;2,6 0 3 3 0

2;2,22 0 9 3 0

2;3,8 0 5 0 0

2;3,24 0 20 2 0

2;4,9 0 11 0 0

2;4,24 1 33 5 0

2;5,7 0 12 1 0

2;5,23 2 62 6 0

2;6,12 0 17 4 1

2;6,22 0 13 7 2

2;7,5 0 50 2 5

2;7,18 0 47 17 0

2;8,2 0 81 27 7

2;8,16 0 62 26 1

2;9,0 0 37 15 6

2;9,14 0 43 7 3

2;9,28 0 47 10 0

2;10,25 0 51 14 1

2;11,1 0 20 11 0

2;11,18 0 25 12 0

3;0,1 0 25 23 0

3;0,15 1 51 40 0

TOTAL 4 748 235 26

Table 13: COL’s Aspectual Production Table 14: ALA’s Aspectual Production

COL


1;8,17 0 0 0 0

1;8,27 0 0 0 0

1;9,17 0 0 0 0

1;9,28 0 2 0 0

1;10,8 0 3 0 0

1;11,1 0 6 1 0

1;11,12 0 4 0 0

1;11,28 0 10 0 0

2;0,12 0 8 2 0

2;0,28 1 10 4 2

2;1,14 0 7 0 0

2;2,0 0 3 0 0

2;2,16 0 16 3 0

2;3,1 0 13 1 0

2;3,16 0 6 4 0

2;3,30 0 15 1 0

2;4,15 0 22 10 0

2;5,0 0 25 5 0

2;5,14 0 9 0 0

2;5,27 0 9 3 0

2;6,10 0 5 3 0

2;6,25 0 50 6 0

2;7,8 0 27 10 0

2;7,22 0 12 3 0

2;8,6 1 18 4 0

2;8,20 0 40 17 0

2;9,11 0 11 9 1

2;9,24 0 6 2 0

2;10,10 0 21 29 0

2;10,21 0 18 7 0

2;11,7 0 19 9 0

TOTAL 2 395 133 3


104

TYA


1;11,25 0 0 0 0

2;0,4 0 0 0 0

2;0,25 0 0 0 0

2;1,5 0 1 0 0

2;1,16 0 0 0 0

2;2,9 0 0 0 0

2;2,20 0 0 0 0

2;3,5 0 0 0 0

2;3,20 0 0 0 0

2;4,6 0 0 0 0

2;4,22 0 0 0 0

2;5,8 0 0 0 0

2;5,24 0 0 1 0

2;6,9 0 0 0 0

2;6,24 0 2 0 0

2;7,7 0 0 0 0

2;7,23 1 1 1 0

2;8,8 1 21 0 0

2;8,22 0 7 2 0

2;9,5 0 8 0 0

2;9,18 0 1 0 0

2;10,2 0 14 3 0

2;10,16 0 12 3 0

2;11,0 0 18 5 0

2;11,14 0 16 7 0

2;11,28 0 25 3 0

3;0,19 0 1 0 0

3;1,1 0 16 65 0

3;1,15 4 11 69 0

3;2,1 0 36 12 0

3;2,15 0 18 20 0

TOTAL 6 208 191 0

Table 15: RJ’s Aspectual Production Table 16: TYA’s Aspectual Production

RJU


1;10,4 0 0 0 0

1;10,14 0 0 0 0

1;11,4 0 3 0 0

1;11,15 0 0 1 0

1;11,26 0 1 1 0

2;0,19 0 3 0 0

2;0,30 0 3 3 0

2;1,15 0 18 2 0

2;2,0 0 11 31 0

2;2,16 0 11 8 0

2;3,1 0 29 21 0

2;3,18 0 14 2 0

2;4,4 0 25 4 0

2;4,19 0 14 11 1

2;5,3 1 18 7 0

2;5,17 0 17 15 0

2;6,2 0 36 5 0

2;6,18 0 12 11 0

2;7,5 0 84 26 1

2;7,15 0 26 29 2

2;7,28 0 19 14 0

2;8,12 0 38 18 2

2;8,26 0 33 8 0

2;9,10 1 24 18 0

2;9,24 1 17 14 0

2;10,7 0 21 21 0

2;10,29 1 11 7 0

2;11,11 1 19 9 0

2;11,25 1 23 50 0

3;0,11 0 35 15 0

3;0,25 0 31 5 0

TOTAL 6 596 356 6


105

SHU


2;1,23 0 7 1 0

2;2,2 0 13 7 0

2;2,11 0 10 7 0

2;3,3 0 18 1 0

2;3,14 0 18 5 0

2;4,7 0 3 1 0

2;4,18 0 4 3 0

2;5,3 0 2 0 0

2;5,18 0 18 2 0

2;6,4 8 21 13 0

2;6,20 0 12 6 0

2;7,16 0 11 3 0

2;7,22 0 11 6 0

2;8,7 1 12 8 0

2;8,22 1 16 6 0

2;9,5 1 9 5 0

2;9,21 0 19 26 1

2;10,6 1 13 9 1

2;10,20 0 3 1 0

2;11,3 1 24 8 1

2;11,16 0 16 8 0

3;0,0 5 36 12 0

3;0,19 2 38 12 2

3;0,28 0 23 37 0

3;1,12 0 24 11 1

3;1,26 0 55 13 2

3;2,17 0 16 13 0

3;2,30 1 20 10 0

3;3,16 0 11 2 1

3;3,27 0 31 28 2

3;4,13 0 20 19 0

TOTAL 21 534 283 11

Table 17: KEM’s Aspectual Production Table 18: SHU’s Aspectual Production

Examining the cumulative data (Table 19), we see where the progressive marker was first

produced at 1;9.5 and the prospective at 1;10.5. Both the retrospective and completive

markers were produced shortly after at 2;0.5.

KEM


2;0,21 0 6 0 0

2;1,0 0 1 0 0

2;1,21 0 1 0 0

2;2,1 0 2 0 0

2;2,12 0 0 0 0

2;3,5 0 0 0 0

2;3,16 0 1 0 0

2;4,1 0 6 0 0

2;4,16 0 3 0 0

2;5,2 0 1 0 0

2;5,18 0 3 1 0

2;6,4 0 3 3 0

2;6,20 1 19 5 0

2;7,5 0 27 35 0

2;7,20 0 41 29 0

2;8,3 0 42 18 0

2;8,19 0 41 27 0

2;9,4 0 40 50 0

2;9,18 0 41 22 0

2;10,1 0 15 9 0

2;10,14 0 38 25 0

2;10,29 0 61 47 0

2;11,12 0 35 24 0

2;11,26 0 69 37 0

3;0,10 1 62 19 0

3;0,24 0 60 61 0

3;1,15 0 69 67 1

3;1,28 0 76 56 0

3;2,11 1 80 69 0

3;2,25 1 79 68 0

3;3,11 0 59 31 0

TOTAL 4 981 703 1


106

ASPECT

AGE (Y;M) COMP PROG PROS RETRO

1;8.0 0 0 0 0

1;8.5 0 0 0 0

1;9.0 0 0 0 0

1;9.5 0 2 0 0

1;10.0 0 4 0 0

1;10.5 0 7 1 0

1;11.0 0 13 0 0

1;11.5 0 11 1 0

2;0.0 0 13 3 0

2;0.5 1 22 4 2

2;1.0 0 15 3 0

2;1.5 0 34 3 0

2;2.0 0 45 44 0

2;2.5 0 43 19 0

2;3.0 0 58 26 0

2;3.5 0 68 10 0

2;4.0 0 67 15 0

2;4.5 1 79 24 1

2;5.0 1 42 8 0

2;5.5 2 109 28 0

2;6.0 8 82 28 1

2;6.5 1 108 35 2

2;7.0 0 199 76 6

2;7.5 1 138 85 2

2;8.0 3 193 71 7

2;8.5 1 204 96 3

2;9.0 1 138 87 7

2;9.5 1 134 75 4

2;10.0 2 127 74 1

2;10.5 0 143 71 1

2;11.0 2 153 87 1

2;11.5 1 111 60 0

3;0.0 6 178 125 0

3;0.5 4 187 86 2

3;1.0 0 130 168 0

3;1.5 4 104 147 2

3;2.0 0 167 81 2

3;2.5 1 114 102 0

3;3.0 2 99 78 0

3;3.5 0 70 33 1

3;4.0 0 31 28 2

3;4.5 0 20 19 0

TOTAL 43 3462 1901 47

Table 19: Total production of Aspect


107

The progressive aspect represents 63% (3462 of 5453 utterances) of the total number of overt

aspectual markers produced followed by 35% (1901) prospective utterances. The production

of the completive and retrospective markers (43 and 47 utterances respectively) together

represents only 2% of the aspectual expressions in the corpus.

Again in order not to rely solely on the frequency of use of the markers, we conducted a

qualitative analysis of the variation in use of the markers. In table 20 and 21 we present the

age at which the markers are used productively with 2 and 5 different predicates respectively.

Using both thresholds, the data reveals that the progressive aspect is the first marker to be

used productively by all participants, followed by the prospective. Examining the data where

the criterion for productive use is 2 predicates, we see that all informants, except for RJU,

used the completive aspectual marker productively before the retrospective marker. This

however does not hold for the 5 predicates criterion as only KEM and SHU demonstrated

productive use of the completive aspect with 5 different verbs. On the other hand, the data

reveals that ALA used the retrospective aspect productively in the 5 predicate criterion

without demonstrating productive use of the completive aspectual marker. Regardless of the

criterion employed, the retrospective aspects are yet to be used productively by TYA and

KEM.

ASPECT COL ALA RJU TYA KEM SHU

RETRO 2;9,11 2;6,22 2;7,5 - - 2;10,6

PROG 1;11,1 1;10,25 1;11,4 2;6,24 2;0,21 2;1,23

PROS 2;0,12 2;2,6 1;11,26 2;7,23 2;6,4 2;2,2

COMP 2;8,6 2;5,23 2;9,10 3;1,15 3;0,10 2;9,5

Table 20: Use of Aspectual markers with 2 different predicates


108

ASPECT COL ALA RJU TYA KEM SHU

RETRO - 2;8,1 - - - 3;1,12

PROG 1;11,12 2;1,5 2;0,19 2;8,8 2;1,21 2;2,2

PROS 2;0,28 2;3,24 2;1,15 2;10,2 2;6,20 2;2,11

COMP - - 2;11,11 - - 3;0,0

Table 21: Use of Aspectual markers with 5 different predicates

Closer analysis of the data reveals that children do not incorrectly overextend progressive

markings to stative verbs (in line with Bickerton (1981), Shirai and Andersen (1995), Erbaugh

(1992), among others). From the 3462 occurrences of the progressive marker, only 3

examples were detected where the progressive was incorrectly used. Note however that all 3

cases are errors involving the marker of negation plus the progressive marker, instead of using

the bare negative marker.

46) * Yo naa niem no Tamir, yo niem Kiisha. (ALA 2;08)

2SG NEG~PROG name NEG Tamir 2SG name Keisha

“Your name is not Tamir, your name is Keisha.”

47) * A naa lov im, a noo lov Ø. (COL 2;07)

1SG NEG~PROG love 3SG 1SG NEG love Ø

“I don’t love him; I don’t love (him).”

48) * Im naa ha noo tiit. (COL 2;09)

3SG NEG~PROG have NEG teeth

“He doesn’t have any teeth.”


109

Additionally, we see where the progressive marker is used target-consistently with the

following arguably stative verbs (verbs of bodily sensation) but with a dynamic reading: pien

‘pain’, fiil ‘feel’, si ‘see’, luk ‘look’ and ier ‘hear’. Closer analysis of the data reveals that

errors of commission are virtually unattested, while errors of omission are robust. There are

only 9 utterances where the progressive marker was incorrectly extended to a prospective

utterance or vice versa, as in (49).

49) *Man a paas dier so. (ALA, 2;05)

Man PROG pass there LOC

“The man is going to pass there.”

We will now examine in detail the errors of omission in the dataset.

6.5.3 Omission of Aspect Markers

We should point out that while the omission of the progressive and prospective markers yields

a difference in temporal interpretation, the completive and retrospective markers may be

freely omitted in JC without changing the primary temporal meaning of the utterance, as

exemplified in (50) – (53).

50) mi a luk fi wan mongki. (TYA 2;10)

1SG PROG look for Q:indef monkey

“I am looking for a monkey.”

Aspect omitted: mi Ø luk fi wan mongki.


110

1SG Ø look for Q:indef monkey

“I looked for a monkey.”

51) Mi a_go yuuz i sopm yaso. (KEM 2;07)

1SG PROS use DET something LOC

“I am going to use the thing here.”

Aspect omitted: mi Ø yuuz i sopm yaso.

1SG Ø use DET something LOC

“I used the thing here.”

52) Manski don iit fi ar aredi. (RJU 3;00)

Manski COMP eat POSS 3SG already

“Manski (completely) ate hers already.”

Aspect omitted: Manski Ø iit fi ar aredi.

Manski Ø eat POSS 3SG already

“Manski ate hers already.”

53) Mi jos bied out a grampaa. (ALA 2;06)

1SG RETRO bathe out PREP grandpa

“I just bathed out by grandpa’s (house).”

Aspect omitted: Mi Ø bied out a grampaa.

1SG Ø bathe out PREP grandpa


111

“I bathed out by grandpa’s (house).”

The difference in interpretation yielded by the presence or absence of the progressive and

prospective markers is subtle, however based on the context of utterance, we are able to

determine where they should have been present and were omitted. Such an analysis allows us

to map more clearly the development of the aspectual zone. We present in Table 22 a detailed

analysis of the development of the progressive and prospective aspect, mapping the

production of the overt progressive and prospective aspectual markers compared to their

respective omissions as deduced from the context of the utterance.


112

Table 22: Production of Progressive and Prospective Aspect

The data shows that children were more likely to produce progressive utterances, whether

overtly marked or not. Additionally, we see that the prospective aspect marker was more

AGE

(Y;M) TOTAL TOTAL

1;8.0 1 1 100 0 0 0 0 0 0 0

1;8.5 2 2 100 0 0 0 0 0 0 0

1;9.0 0 0 0 0 0 1 1 100 0 0

1;9.5 22 20 91 2 9 5 5 100 0 0

1;10.0 15 12 80 3 20 7 7 100 0 0

1;10.5 27 20 74 7 26 11 10 91 1 9

1;11.0 57 44 77 13 23 19 19 100 0 0

1;11.5 37 26 70 11 30 22 21 95 1 5

2;0.0 43 31 72 12 28 12 9 75 3 25

2;0.5 69 47 68 22 32 30 26 87 4 13

2;1.0 86 71 83 15 17 40 37 93 3 8

2;1.5 111 77 69 34 31 45 42 93 3 7

2;2.0 111 66 59 45 41 99 55 56 44 44

2;2.5 121 78 64 43 36 73 56 77 17 23

2;3.0 105 47 45 58 55 87 61 70 26 30

2;3.5 151 83 55 68 45 57 47 82 10 18

2;4.0 119 54 45 65 55 49 34 69 15 31

2;4.5 138 59 43 79 57 73 49 67 24 33

2;5.0 87 45 52 42 48 52 44 85 8 15

2;5.5 174 65 37 109 63 98 70 71 28 29

2;6.0 114 32 28 82 72 64 36 56 28 44

2;6.5 142 36 25 106 75 74 39 53 35 47

2;7.0 239 38 16 201 84 107 32 30 75 70

2;7.5 178 43 24 135 76 123 38 31 85 69

2;8.0 228 42 18 186 82 121 50 41 71 59

2;8.5 223 25 11 198 89 118 22 19 96 81

2;9.0 170 30 18 140 82 112 25 22 87 78

2;9.5 158 29 18 129 82 120 46 38 74 62

2;10.0 152 29 19 123 81 93 19 20 74 80

2;10.5 157 15 10 142 90 93 22 24 71 76

2;11.0 175 24 14 151 86 117 30 26 87 74

2;11.5 131 21 16 110 84 78 18 23 60 77

3;0.0 195 17 9 178 91 154 29 19 125 81

3;0.5 211 24 11 187 89 98 12 12 86 88

3;1.0 147 17 12 130 88 194 26 13 168 87

3;1.5 113 9 8 104 92 167 20 12 147 88

3;2.0 179 11 6 168 94 100 19 19 81 81

3;2.5 126 12 10 114 90 113 11 10 102 90

3;3.0 103 4 4 99 96 89 11 12 78 88

3;3.5 76 6 8 70 92 35 2 6 33 94

3;4.0 32 1 3 31 97 30 2 7 28 93

3;4.5 20 0 0 20 100 27 8 30 19 70

TOTAL 4745 1313 28 3432 72 3007 1110 37 1897 63

PROSPECTIVE ASPECT

0 PROS % PROS %0 PROG % PROG %

PROGRESSIVE ASPECT


113

likely to be omitted (37% omission) when compared to the progressive marker (28%

omission).

For a more graphic analysis, we present in Figure 2, the percentage of production of the overt

markers. We see a gradual increase in the use of the overt markers for both the progressive

and prospective aspects. While the prospective marker was produced only a month following

the production of the progressive marker, its development appears to be slower than that of

the progressive aspect. At 2;3 the progressive marker was used in over 50% of the progressive

utterances; however the use of the prospective marker did not reach the 50% bar until 4

months later, at 2;7.

Figure 2: Percentage production of Overt Progressive and Prospective Aspect in utterances

expressing progressive or prospective aspectual readings.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%PROGRESSIVE

PROSPECTIVEPRODUCTION

AGE (Y;M)


114

(54) - (59) are examples of utterances with aspectual interpretations, but for which the

grammatical markers are omitted. These are child-specific constructs that, based on the

contexts of utterance and intended interpretation, are clearly omitting the overt aspectual

distinctions, and as such are not in line with adult JC grammar.

54) Ø Ø uol mi fut. (SHU 2;03)

Ø Ø hold 1SG foot

“She is holding my foot.”

55) Mi Ø fiks i bak. (ALA 2;02)

1SG Ø fix 3SG back

“I am fixing it back.”

56) Ø Ø bons it. (COL 1;11)

Ø Ø bounce 3SG

“I am bouncing it.”

57) Felisha Ø kyari mi. (RJU 2;05)

Felisha Ø carry 1SG

“Felisha is going to carry me.”

58) Ø Ø jrap aaf. (KEM 2;00)

Ø Ø drop off

“They are going to fall off.”


115

59) Ø Ø shuo Moesha i kyat. (TYA 2;08)

Ø Ø show Moesha DET cat

“I am going to show Moesha the cat.”

Recall that in adult JC, bare non-stative verbs are interpreted as past tense; however as

demonstrated by the above examples, children employ bare verbs also in contexts where there

are present progressive (54-56) and prospective (57-59) interpretations. These utterances,

which express an aspectual reading but remain unmarked, appear to be evidence that JC

involves a Root Infinitive (RI) stage. This aspectual conditioning of bare verbs is also attested

in Capeverdean in support of the RI hypothesis (Pratas and Hyams, 2009). We will examine

this phenomenon in more detail in Section 6.6.

6.5.4 Intermediate Conclusion

In this section we saw where the overt progressive aspect marker was the first to be produced

and used productively followed closely by the prospective aspect, and then by the completive

marker (in accordance with the 2 predicates criterion). Despite the prospective marker being

produced just a month after the progressive, its development and use tends to be much slower:

not until 2;7 did the prospective marker reach 50% production while at 2;3 months the

progressive marker was used in over 50% of all utterances expressing progressive aspect. The

completive and retrospective markers both came on stream at 2;0.5 but was not used

productively until much later. Overt aspectual markers are therefore attested early in JC in the

sequence: AspProgressive > AspProspective> AspCompletive> AspRetrospective. As the children develop,

we see an influx in the production of the overt progressive and prospective markers.


116

In returning to the ‘Aspect before Tense hypothesis’, we observed that Progressive,

Prospective, Completive and Retrospective Aspect were all attested in the corpus before 2;1,

the point where the first overt Past Tense morpheme was attested. Using this as a benchmark,

the data would seemingly support the ‘Aspect before tense hypothesis’. However given that

we are unable to provide conclusive evidence as to the point of acquisition of past tense (due

to the past interpretation associated with bare non-stative verbs), it would be safer to conclude

that aspect is overtly expressed before tense is overtly expressed in children acquiring JC. The

use of bare verbs to express an aspectual interpretation appears to be evidence of root

infinitives and will be the subject of the next section.

6.6 ROOT INFINITIVES

In this section, we examine the phenomenon of Root Infinitives (RIs). RIs are main clauses

containing uninflected verbal forms which often (though not always) allow subject drop

(Rasetti 2000). According to Hamann and Plunkett (1998) RIs are linked to null subjects as

both phenomena decline at roughly the same time. They are expected in languages that are

non-pro-drop and/or languages in which the non-finite verb forms do not raise higher than T

in the inflectional space (Rizzi 1993/1994). According to Pratas and Hyams (2009)

Capeverdean (a Portuguese-based creole) children produced null subjects in root sentences at

proportions roughly equivalent to children acquiring other non-pro-drop languages: between

28 – 35 months 37% and between 36 to 42 months 17%. Additionally, based on results from

an experimental investigation of finiteness, they concluded that Capeverdean children treat

bare verbs differently from adults: while in the target language bare verbs are interpreted as

past tense, for children they are interpreted as progressive aspect. This aspectual conditioning

of bare verbs supports the hypothesis of an RI stage.


117

We will now examine this phenomenon in Jamaican Creole

6.6.1 Root Infinitives in JC

This section is devoted to the analysis of bare root infinitival verbs in the utterances of

children acquiring JC. Recall that in adult JC, bare non-stative verbs are interpreted as having

a past reading. However as demonstrated in the previous section, and further shown by

examples (60) – (71), children behave very differently from adults in that all bare forms do

not pattern with the bare form in the target language in yielding a past reading. Children

employ bare verbs also in contexts where there is an aspectual interpretation thereby

providing evidence for the hypothesis that JC involves an RI stage.

60) Ø Ø chobl i maabl. (ALA 2;04)

Ø Ø trouble DEF marble

“He is troubling the marble.”

61) Ø Ø bons it. (COL 1;11)

Ø Ø bounce 3SG

“I am bouncing it.”

62) Ø Ø chuo oot i. (KEM 2;00)

Ø Ø throw out it

“I am throwing it out.”

63) Ø Ø jraiv it. (RJU 1;11)


118

Ø Ø drive 3SG

“She is driving it.”

64) Ø Ø uol mi fut. (SHU 2;03)

Ø Ø hold 1SG foot

“She is holding my foot.”

65) Ø Ø riid buk. (TYA 2;07)

Ø Ø read book

“They are reading a book.”

66) Ø Ø gi guoti. (ALA 1;10)

Ø Ø give goatie

“She is going to give goatie.”

67) Ø Ø bied im outsaid. (COL 1;11)

Ø Ø bathe 3SG outside

“He is going to bathe him outside.”

68) Ø Ø jrap aaf. (KEM 2;00)

Ø Ø drop off

“They are going to fall off.”

69) Ø Ø torn aan i. (RJU 2;01)

Ø Ø turn on 3SG


119

“I am going to turn it on.”

70) Ø Ø bring som priti blous. (SHU 2;05)

Ø Ø bring some pretty blouse

“She is going to bring some pretty blouses.”

71) Ø Ø shuo Moesha i kyat. (TYA 2;08)

Ø Ø show Moesha DET cat

“I am going to show Moesha the cat.”

Also we see where RIs are produced in utterances where the subject is overtly expressed, as in

(72) – (77). This is however attested for only 16% (395 of 2424 RIs utterances) of the data. So

RIs are much more frequent in environments where the subject is dropped (84%), compared

to environments where the subject is overtly pronounced.

72) Mi Ø fiks i bak. (ALA 2;02)

1SG Ø fix 3SG back

“I am fixing it back.”

73) Jan Ø daans. (COL 1;10)

John Ø dance

“John is dancing.”

74) Mi Ø ron gaan lef yo. (KEM 3;03)

1SG Ø run gone leave 2SG


120

“I am running leave you.”

75) Felisha Ø kyari mi. (RJU 2;05)

Felisha Ø carry 1SG

“Felisha is going to carry me.”

76) Yu mada Ø kom fi yu an biit yu. (SHU 2;09)

2SG mother Ø come for 2SG and beat 2SG

“Your mother is going to come for you and beat you.”

77) Mii an Aleks Ø kola oova de-so. (TYA 3;01)

1SG and Alex Ø colour over LOC

“Alex and I are going to colour over there.”

Further analysis of the data reveals that null subjects are much more frequent with RIs than

with finite clauses14

. As demonstrated in Table 23 below, the total proportion of null subjects

in finite declarative clauses15

in JC is a mere 23.2% when compared to 88% null subjects with

RIs. As detailed in the table, this is in line with findings from other languages. The

comparable proportion also supports the hypothesis that child JC has RIs.

14

It is not very clear how significant the finite/nonfinite distinction is in creole languages (Mufwene 1999; Dijkhof & Mufwene 1989). For this analysis, only infinitives are analyzed as non-finite. 15 Only non-stative declarative utterances were involved in this analysis.


121

Null subj

+fin

% Null subj

RI

%

JC

COL

ALA

RJU

TYA

KEM

SHU

447/1502

397/1923

467/1638

185/747

575/2502

176/1422

29.8

20.6

28.5

24.8

23.0

12.4

380/449

317/364

329/400

224/240

564/599

154/173

84.6

87.1

82.3

93.3

94.2

89.0

French

Daniel (Pierce 1989)

Nathalie (Pierce 1989)

Philippe (Pierce 1989)

Augustin (Rasetti 1999)

Marie (Rasetti 1999)

150/273

90/304

182/782

157/582

154/560

54.9

29.6

23.3

26.8

27.5

166/205

131/295

153/194

66/71

130/134

81.0

44.4

78.9

93.0

97.0

German

Simone (Behrens 1993)

Andreas (Krämer 1993)

781/3699

34/263

21.1

12.9

2199/2477

69/101

88.8

68.3

Dutch

Thomas (Krämer 1993)

Heinz (Haegeman 1995a)

165/596

1199/3768

27.7

31.8

246/267

615/721

92.1

85.3

Flemish

Maarten (Krämer 1993)

23/62

25.0

89/100

89.0

Faroese

O. (Jonas 1995)

8/52

15.4

67/161

41.6

Danish

Anne (Hamann & Plunkett 1998)

Jens (Hamann & Plunkett 1998)

366/3379

742/3173

10.8

23.4

394/667

539/937

59.1

57.5

Table 23: Subject Omission in Finite and Root Infinitival clauses16

Figure 3 and corresponding Table 24 demonstrate the production of RIs with progressive and

prospective aspectual interpretations in the corpus.

16 Data for all languages except JC is from Rasetti (2000).


122

Figure 3: Production of RIs with aspectual interpretations

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

% P

RO

DU

CTI

ON

AGE (Y;M)


123

AGE (mths) Omitted Aspect Overt Aspect Total Aspect % Root infinitives

1;9.5 25 2 27 93

1;10.0 19 3 22 86

1;10.5 30 8 38 79

1;11.0 63 13 76 83

1;11.5 47 12 59 80

2;0.0 40 15 55 73

2;0.5 73 26 99 74

2;1.0 108 18 126 86

2;1.5 119 37 156 76

2;2.0 121 89 210 58

2;2.5 134 60 194 69

2;3.0 108 84 192 56

2;3.5 130 78 208 63

2;4.0 88 80 168 52

2;4.5 108 103 211 51

2;5.0 89 50 139 64

2;5.5 135 137 272 50

2;6.0 68 110 178 38

2;6.5 75 141 216 35

2;7.0 70 276 346 20

2;7.5 81 220 301 27

2;8.0 92 257 349 26

2;8.5 47 294 341 14

2;9.0 55 227 282 20

2;9.5 75 203 278 27

2;10.0 48 197 245 20

2;10.5 37 213 250 15

2;11.0 54 238 292 18

2;11.5 39 170 209 19

3;0.0 46 303 349 13

3;0.5 36 273 309 12

3;1.0 43 298 341 13

3;1.5 29 251 280 10

3;2.0 30 249 279 11

3;2.5 23 216 239 10

3;3.0 15 177 192 8

3;3.5 8 103 111 7

3;4.0 3 59 62 5

3;4.5 8 39 47 17

Table 24: Production of RIs with Aspectual Interpretations

As Pratas & Hyams (2009) pointed out for Capeverdean, two main reasons that lead us to

expect an RI stage in JC are: i) it is a non-null subject language; and ii) it has no agreement

morphology. Both of these factors are predictive of languages which have an RI stage (Sano


124

& Hyams 1994; Wexler 1994; Rizzi 1993/1994). Rizzi (1993/1994) argues that RIs are

truncated structures that arise as a consequence of the option available to the child to ‘strip

off’ external clausal layers. This option is also responsible for the property of null subjects in

early grammars. Various researchers (such as Hamann & Plunkett 1998; Rizzi 2000; among

others) shows where there is a striking correspondence between the decline of both

phenomena roughly around the same period. We will examine the production of null subjects

in more detail in the next chapter.

One empirical observation is that RIs seem to be prevented in wh-questions. For Dutch,

Haegeman (1995a) found 16% RI among non-interrogative clauses and 2.5% among wh-

interrogatives. Similarly, for French, Crisma (1992) reported 15% nonfinite verbs in

declaratives and 0% in wh-questions. The current finding for JC reveals a striking asymmetry.

As demonstrated in Table 25 below, the proportion of root infinitival wh-interrogatives is

much lower than that produced in declaratives. While there are 18.5% RIs among

declaratives, only 2% RIs (22 of 1099 cases) was produced in wh-interrogatives17

. If the

proportion had been constant across clauses, we would expect to find around 203 instances of

interrogative RIs.

Informants RI in

declaratives

% RI in

wh-interrogatives

%

COL

ALA

RJU

TYA

KEM

SHU

449/1951

364/2287

400/2038

240/987

599/3101

154/1576

23.0

15.9

19.6

24.3

19.3

9.7

2/131

1/212

10/181

0/36

2/101

7/438

1.5

0.5

5.5

0.0

2.0

1.6

Total 2206/11940 18.5 22/1099 2.0

Table 25: RIs with Finite Declaratives and Wh-Interrogatives

17

For a coherent comparison, like for declarative clauses, only non-stative interrogative utterances were included in this analysis.


125

These data clearly indicate that there is a structural difference between finite utterances and

RIs whereby the latter are excluded from contexts requiring the projection of the CP, whether

it is overtly realized or null. As revealed, of the 22 cases of interrogative RIs, 9 were produced

in wh-constituent utterances for which the wh-marker was omitted compared to 13 produced

in wh-questions with overt wh-elements. This is exemplified in (78) and (79) respectively.

The production of null and overt wh-questions will be discussed in the following chapter.

78) Ø dadi go? (COL 2;05)

Ø Daddy go

“Where is daddy going?”

79) We im du? (RJU 2;10)

What 3SG do

“What is he doing?”

We will now turn to a discussion of the overall development of TMA in JC.

6.7 CUMULATIVE DEVELOPMENT OF THE TMA ZONE

Although there is significant individual variation with the age of attestation of the various

markers as demonstrated in the previous sections, the children show many similarities in the

qualitative development of the TMA zone. To capture this developmental pattern we divided

the corpus, somewhat arbitrarily into phases on the basis of MLU18

as discussed in the

18

The MLU is word based: both lexical and functional items are treated as individual words. The MLU for JC therefore differs from languages with morphologically complex words.


126

previous chapters. For this analysis, we grouped the single word stage (MLU <1.49) and the

two word combination stage (MLU 1.5 – 2.49) to form Phase 1 which corresponds to MLU

under 2.5. For Phase 2, we presented utterances produced at MLU 2.5 - 3.49, and Phase 3

includes more complex combinations where MLU is 3.5 and over. Table (26) details Phase

119

.

PHASE 1

INFORMANT COL ALA RJU TYA KEM SHU TOTAL

AGE RANGE 1;8,17 - 1;11,28

1;9,25 - 2;1,20

1;10,4- 2;0,30

1;11,25 - 2;7,23

2;0,21 - 2;4,16 -

1;8,17 - 2;7,23

ASP COMP 0 0 0 1 0 - 1

ASP PROS 1 0 5 2 0 - 8

ASP PROG 25 24 10 4 20 - 83

ASP RETRO 0 0 0 0 0 - 0

MOD ABL 0 0 2 0 1 - 3

MOD OBL 0 0 0 0 0 - 0

MOD NEC 0 0 0 0 0 - 0

MOD EPIS 0 0 0 0 0 - 0

MOD WI 0 0 0 0 0 - 0

PAST TENSE 0 0 0 0 0 - 0

TOTAL 26 24 17 7 21 0 95

Table 26: TMA Phase 1

In phase 1 the main TMA marker used is the progressive (also attested in English, as

demonstrated by Boland 2006). It is already used rather frequently, with the exception of

TYA, who used it only 4 times (4.8% of the progressive aspectual marker in the corpus at this

stage). The prospective aspect and ability modal is also used very sporadically. Only TYA has

a single occurrence of the Completive marker in phase 1. The children however show a sharp

increase in the number of TMA markers in phase 2 as detailed in Table 27 below. RJU’s

development is most extreme, showing an increase from 17 TMA expressions to 354. KEM’s

development is much slower, moving from a total of 21 TMA expressions to only 38. Closer

19 No data is included for SHU at this phase as at the start of the recordings her MLU was already over 2.5.


127

analysis of KEM’s files reveals however that he did not remain at the second phase for very

long, as after 4 recordings at this stage he has quickly advanced to an MLU over 3.5.

PHASE 2


AGE RANGE 2;0,12 - 2;6,25

2;2,6 - 2;5,7

2;1,15 - 2;6,18

2;8,8 - 2;10,16

2;5,2 - 2;6,20

2;1,23 - 2;6,20

2;0,12 - 2;10,16

ASP COMP 1 1 1 1 1 8 13

ASP PROS 42 14 117 8 9 46 236

ASP PROG 198 93 205 63 26 126 711

ASP RETRO 2 0 1 0 0 0 3

MOD ABL 12 14 24 1 2 14 67

MOD OBL 1 1 0 0 0 1 3

MOD NEC 0 0 0 0 0 0 0

MOD EPIS 0 0 1 0 0 0 1

MOD WI 6 1 4 0 0 1 12

PAST TENSE 2 0 1 0 0 0 3

TOTAL 264 124 354 73 38 196 1049


In phase 2, we note a significant increase in the use of the progressive markers, moving from

83 occurrences to 711. All children are now using the prospective markers; RJU being

exceptionally advanced. The completive aspect is attested once in all the children’s data, with

the exception of SHU who produced it 8 times. Closer analysis of the 8 occurrences of the

completive aspect marker in SHU’s data reveals that it was not being used productively as it

was restricted to a single verb in only one file. Also we see where the ability modal is used

frequently by all children, with the exception of TYA and KEM who produced it only once

and twice respectively. A new marker that is produced in this stage is the modal wi. It is

sporadically produced by all children with the exception of TYA and KEM. The retrospective

aspect, obligation modal, epistemic modal and past tense markers were sporadically produced

while the necessity modal remained unattested.


128

In phase 3, MLU is now over 3.5 and there is a considerable increase in the use of TMA

expressions, moving from a total of 1049 in phase 2 to 5292. The progressive aspect is still

the marker that is predominately used, followed by the prospective aspect and the ability

modal. We see a small increase in use of the other TMA markers however with much

individual variation. While all markers are produced at least once in each child’s production,

TYA is yet to produce a single retrospective aspect, epistemic modal or a past tense marker.

PHASE 3


AGE RANGE 2;7,8 - 2;11,7

2;5,23 - 3;0,15

2;7,5 - 3;0,25

2;11,0 - 3;2,15

2;7,5 - 3;3,11

2;7,16 - 3;4,13

2;5,23-3;4,13

ASP COMP 1 3 5 4 3 13 29

ASP PROS 90 221 381 181 694 237 1804

ASP PROG 172 631 234 141 935 408 2521

ASP RETRO 1 26 5 0 1 11 44

MOD ABL 109 187 98 56 124 113 687

MOD OBL 0 42 2 2 11 15 72

MOD NEC 1 5 3 2 1 12 24

MOD EPIS 0 29 0 0 2 1 32

MOD WI 1 8 10 1 27 2 49

PAST TENSE 1 6 3 0 14 6 30

TOTAL 376 1158 741 387 1812 818 5292


6.7.1 Co-occurrence of TMA Markers

In examining the cumulative development of the TMA zone in JC we note that there are

combinations of markers in the same utterance. It is our aim to study the sequence of the co-

occurrence of the markers in order to account for the cartographic development20

of the IP.

We have already observed in Section 6.4.2 that the progressive marker is seen to co-occur

20 See Shlonsky (2010) for overview of Cartography within syntactic theory.


129

with the overt past tense marker yielding a past progressive interpretation as exemplified in

(80) – (84):

80) Mii did a fiid im. (ALA 2;07)

1SG PAST PROG feed 3SG

“I was feeding him.”

81) Mi did a sliip. (COL 2;01)

1SG PAST PROG sleep

“I was sleeping.”

82) Yo ben a jraiv i van? (KEM 2;11)

2SG PAST PROG drive DET van

“Were you driving the van?”

83) We im daa go? (RJU 2;07)

where 3SG PAST~PROG go

“Where was he going?”

84) Mi wehn a ron wid Lietn an a jap. (SHU 2;11)

1SG PAST PROG run with Leighton and 1SG drop

“I was running with Leighton and I fell.”

Table 29 provides detail of the individual production of the co-occurrence of the past tense

marker and progressive aspectual marker.


130

AGE (Y;M) COL ALA RJU TYA KEM SHU

2;1.0 1 0 0 0 0 0

2;1.5 0 0 0 0 0 0

2;2.0 0 0 0 0 0 0

2;2.5 0 0 0 0 0 0

2;3.0 0 0 0 0 0 0

2;3.5 0 0 0 0 0 0

2;4.0 0 0 0 0 0 0

2;4.5 0 0 0 0 0 0

2;5.0 0 0 0 0 0 0

2;5.5 0 0 0 0 0 0

2;6.0 0 1 0 0 0 0

2;6.5 0 0 0 0 0 1

2;7.0 0 6 0 0 0 0

2;7.5 0 1 2 0 0 0

2;8.0 0 0 0 0 0 0

2;8.5 0 2 3 0 0 0

2;9.0 0 2 0 0 0 0

2;9.5 0 0 0 0 0 0

2;10.0 0 0 0 0 0 0

2;10.5 0 0 0 0 0 1

2;11.0 0 0 0 0 1 1

2;11.5 0 1 0 0 0 0

3;0.0 0 0 0 0 0 0

3;0.5 0 0 0 0 0 0

3;1.0 0 0 0 0 2 0

3;1.5 0 0 0 0 1 2

3;2.0 0 0 0 0 0 3

3;2.5 0 0 0 0 0 0

3;3.0 0 0 0 0 0 3

3;3.5 0 0 0 0 2 0

3;4.0 0 0 0 0 0 1

3;4.5 0 0 0 0 0 1

TOTAL 1 13 5 0 6 13

Table 29: Co-occurrence of Overt Tense with Progressive Aspect

We observed that the first utterance where a past tense marker was combined with a

progressive marker is attested in COL’s production at 25 months of age. This is his only

sporadic combination yielding the past progressive. ALA is next in producing this

combination at 30 months of age followed by SHU at approximately 30 and half months, RJU

at 31 and half months and finally KEM at 35 months of age. The combination is yet to be


131

attested in the speech of TYA. The data reveals much individual variation with regards to the

age of production and the frequency of this combination. However it should be noted that the

production of this combination is 100% consistent with the cartographic hierarchy of the

target system, in that the Tense marker is always realized before the Progressive marker. At

no time did the children produce a combination where the Progressive marker was situated

above the Tense marker.

In line with the discussion of phases presented above, Table 30 reveals that in Phase 1 (MLU

< 2.5) no utterances were produced where the past morpheme was combined with the

progressive aspect marker yielding a past-progressive reading while at phase 2 only 2

combinations were attested. We see the increase of these utterances at Phase 3 where MLU is

over 3.5. At this phase the combination of the past with the progressive marker is more

common for most children except for COL, whose only sporadic combination was in Phase 2

and TYA who is yet to start combining markers.

PHASES COL ALA RJU TYA KEM SHU TOTAL

PHASE 1 0 0 0 0 0 0 0

PHASE 2 1 0 0 0 0 1 2

PHASE 3 0 13 5 0 6 12 36

Table 30: Distribution of Co-occurrences of the Past and Progressive Marker

A closer analysis of the data reveals that other combinations of TMA markers were evident in

the children’s production. Past tense was seen to co-occur with ability modal (85); the

necessity modal co-occurred with progressive aspect (86); retrospective aspect co-occurred

with completive aspect (87) and progressive aspect (88); and epistemic modal co-occurred

with the obligation modal (89).


132

85) Im did kyahn waak. (RJU 2;08)

3SG PAST ABL~NEG walk

“He couldn’t walk.”

86) Ø mos a riid. (RJU 2;08)

Ø NEC PROG read

“She must be reading.”

87) Ø jos don bied. (COL 2;00)

Ø RETRO COMP bathe

“I just finished having a bath.”

88) Ø jos a kum bak. (RJU 2;04)

Ø RETRO PROG come back

“He is just coming back.”

89) A wuda afi get som jakit. (ALA, 2;10)

1SG EPIS OBL get some jacket

“I would have to get some jacket.”

As demonstrated in the examples above, the first combination of TMA markers was within

the aspectual zone. At 2;00 the retrospective aspectual marker was combined with the

completive marker in COL’s production, and at 2;04 it was combined with the progressive

marker in RJU’s dataset. Apart from the inter-zonal combinations resulting in the production


133

of the past progressive marker as discussed above, we note the introduction of combinations

of other markers from different TMA zones in the speech of RJU: past tense was combined

with ability modal, and necessity modal was combined with progressive aspect. Both

combinations were produced at 2;08. Another intra-zonal combination was observed in the

modal zone in the speech of ALA: the epistemic modal was combined with the obligatory

modal at 2;10.

Due to the scarcity of co-occurrence of markers in the same phrase within the TMA zone, the

data does not lend itself to a discussion on the relative sequence in which the combinations

were produced. Apart from the co-occurrence of the past and progressive markers, other

combinations are only sporadically attested. Nonetheless, in examining the combinations we

note that they were all target-consistent. Recall the TMA zone in the target is as follows:

Epistemic modal > Past tense; Future tense > Necessity modal > Obligation modal >

Ability/permission modal > Retrospective aspect > Progressive aspect > Prospective aspect

> Completive aspect. The children’s production reveals that the combined markers all

followed the order of the clausal hierarchy attested in the target language. At no time was a

marker which is situated lower in the TMA zone seen to occur before a higher element. The

following hierarchical development was demonstrated:

90) i. Past tense > progressive aspect (80 - 84)

ii. Past tense > ability/permission modal (85)

iii. Necessity modal > progressive aspect (86)

iv. Retrospective aspect > completive aspect (87)

v. Retrospective aspect > progressive aspect (88)

vi. Epestemic modal > obligatory modal (89)


134

The attested sequences outlined in (90) above confirm that children acquiring JC do not

entertain the possibility of target-inconsistent orders in their development of TMA, while the

possibility of other target inconsistencies, such as omissions, are evident. This target-

consistent hierarchical development provides evidence that the child is knowledgeable of the

rules governing combinations from an early age. Structures that are high in the clause are

correctly combined with structures located in a lower domain. The target-consistent

combination of structures attested at different clausal levels call for a full competence

approach to the development of the cartographic sequence.

We now examine the distribution of the markers in comparison to their distribution in the

adult language.

6.7.2 Correlation of Children’s Utterance with Input Data

Based on the minimal attestation of some TMA markers compared to others that are used with

great frequency, in establishing the relative order of acquisition, we have examined the input

data. We have seen where the markers in the input are basically stable across the three phases.

This suggests that there is no ‘fine tuning’ by the adults to match properties of the child

systems. The data also reveal that the use of the markers in the target language is skewed in

the same direction as in the children production. The proportion of progressive and

prospective markers are much higher than other aspectual markers, likewise the

ability/permission modal is used with much greater frequencies than the other categories of

modals. The following proportions, as demonstrated in Figure 4, represent the distribution of

TMA in the input throughout the 3 phases: progressive aspect21

~60%; prospective aspect

21

The total proportion of progressive and prospective aspects is estimated based on actual calculations of a section of the dataset.


135

~16%; ability/permission modal ~12% and all others grouped together ~12%. Detailed

distribution of TMA in the input can be found in Appendix 3.

Figure 4: TMA in the input and in child production

Although there is a high correlation between the distribution of the TMA markers in the input

and in the child production, there are differences. In phase 1 the proportion of progressive

aspects is larger than in the input while the prospective and ability/permission markers are

lower. In phase 2 both the progressive and prospective markers are larger than the input. In

stage 3 there is a great increase in the use of the prospective, a decrease in the use of the

progressive and the proportion of ability/permission modal is equal to that in the input. The

input may therefore have some influence on the distribution but, based on its relative

uniformity across the three phases and the attested differences in the children’s utterance, a

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

CHILD UTTERANCE

INPUT DATA

PHASPHASE 1 PHASE 2 PHASE 3


136

parallel production of child directed speech and child speech does not seem to be supported.

The input therefore, in and of itself, cannot account for the development of TMA markers in

the child production. The main findings will be summarized and a supplementary analysis

accounting for the data will be discussed in the next section.

6.8 DISCUSSION & CONCLUSION

6.8.1 Summary of Main Findings

This work concentrated on exploring the development of the IP domain in early JC and it

focused on the core TMA markers. It was found that the ability/permission modal was the

first modal to be produced and used productively. The obligation, necessity, epistemic, and

prediction/intention modal wi, however, were very sparse in the dataset and individual

variation was significant. We saw that children rarely explore the option of overtly marking

the verb for past tense and as such conclusive evidence could not be established regarding the

development of the overt past tense marker. The progressive and prospective aspect markers

were the first and most robust aspectual markers produced and used productively in the

corpus. The omission of these markers however in contexts with aspectual meaning provides

evidence of a root infinitive stage in early JC. Though TMA combinations are sporadic,

whenever they co-occur within a clause, they are always consistent with the cartographic

sequence of the adult system. Additionally, the distribution of the TMA markers reveals a

striking correlation with the input data: progressive is most robustly attested followed by the

prospective and the ability/permission modal.


137

6.8.2 Discussion

The sequence for which the various overt morphemes were attested in the IP zone and the

order of productive use (based on the 2 predicate criterion) for each participant are presented

in (a) and (b) respectively as follows:

91) COL:

a) Aspprogressive (1;9,28) > Aspprospective (1;11,1) > Aspcompletive and Aspretrospective (2;0,28)>

Tense (2;1,14) > Modability/permission (2;3,16) > Modwi (2;3,30) > Modobligation (2;5,14)

> Modnecessity (2;8,6)

b) Aspprogressive (1;11,1) > Aspprospective (2;0,12) > Modability/permission (2;3,30) > Modwi

(2;4,15) > Tensepast (2;5,0) > Aspcompletive (2;8,6) > Aspretrospective (2;9,11)

92) ALA:

a) Aspprogressive (1;10,4) > Aspprospective (2;2,6) > Modwi (2;2,22) > Modability/permission

(27;08) > Aspcompletive & Modobligation (2;4,24) > Modepestemic (2;5,23) > Aspretrospective

(2;6,12) > Modnecessity and Tense (2;6,22)

b) Aspprogressive (1;10,25) > Aspprospective (2;2,6) > Modability/permission (2;3,8) > Modwi ~

Aspcompletive (2;5,23) > Modepistemic (2;6,5) > Modobligation (2;6,12) > Aspretrospective

(2;6,22) > Modnecessity (2;7,18) > Tensepast (2;11,1)

93) RJU:

a) Aspprogressive (1;11,4) > Aspprospective (1;11,15) > Modability/permission (1;11,26) > Modwi

(2;3,1) > Modepestemic (2;4,4) > Aspretrospective & Tense (2;4,19) > Aspcompletive (2;5,3)

> Modnecessity (2;8,26) > Modobligation (2;11,25)


138


(2;6,18) > Aspretrospective (2;7,5) > Tensepast (2;8,12) > Aspcompletive (2;9,10) > Modobligation

~Modnecessity (3,0,25)

94) TYA:

a) Aspprogressive (2;1,5) > Aspprospective (2;5,24) > Aspcompletive (2;7,23) > Modability/permission

(2;10,2) > Modnecessity (2;11,14) > Modwi (3;1,1) > Modobligation (3;1,15)

b) Aspprogressive (2;6,24) > Aspprospective (2;7,23) > Modability/permission (2;11,0) > Aspcompletive

(3;1,15)

95) KEM:

a) Aspprogressive (2;0,21) > Modability/permission (2;1,21) > Aspprospective (2;4,18) > Aspcompletive

(2;6,20) > Modwi (2;8,3) > Modepestemic (2;9,4) > Modobligation (2;9,18) > Tense (2;10,29)

> Modnecessity (3;0,10) > Aspretrospective (3;1,15)


(2;8,19) > Aspcompletive (3;0,10) > Tensepast (3;0,24) > Modobligation (3;1,15)

96) SHU:

a) Aspprogressive & Aspprospective (2;1,23) > Modability/permission (2;3,14) > Aspcompletive &

Modobligation (2;6,04) > Modwi & Tense (2;6,20) > Aspretrospective (2;9,21) > Modnecessity

(2;11,16) >Modepestemic (3;0,0)

b) Aspprogressive (2;1,23) > Aspprospective (2;2,2) > Modability/permission (2;5,18) > Aspcompletive

~ Tensepast (2;9,5) > Aspretrospective (2;10,6) > Modwi (2;11,3) > Modnecessity (3;0,0) >

Modobligation (3;0,28)


139

The overall sequence for which the various overt morphemes were first attested in the IP zone

is therefore:

97) Aspprogressive (1;09) > Aspprospective (1;10) > Modability/permission (2;00) >

Aspcompletive & Aspretrospective (2;00) > Tense (2;01) > Modwi (2;02) >

Modepistemic (2;04) > Modobligation (2;04) > Modnecessity (2;06)

and the order of productive use (in keeping with the 2 predicate criterion) is:

98) Aspprogressive (1;10,25) > Aspprospective (1;11,26) > Modability/permission (2;0,30) >

Modwi (2;4,15) > Tensepast (2;5,0) > Aspcompletive (2;5,23) > Modepestemic (2;6,5)

> Modobligation (2;6,12) > Aspretrospective (2;6,22) > Modnecessity (2;7,18)

On the basis of frequency and productive use, the hypothesis that structure emerges overtime

incrementally in a bottom-up manner may be too absolute and cannot account for the current

findings. We see that all informants start building the aspectual zone with the following

structure: Aspprogressive > Aspprospective > Modability/permission, but individual variation is evidenced

thereafter.

In the target system only epistemic modals scope above Tense, while all root modals and

aspectual categories scope below it. The difference in distribution between the modal types

with regard to Tense is linked to the difference in semantic interpretation: modals above

Tense are concerned with the speaker’s deductions or opinions while those below Tense are

strictly subject-oriented properties (Cinque 1999). Examining the order of attestation of TMA

in the present corpus, we see that some of the root modals and aspectual markers do not fit in

the predicted categories. Of all the root modals only the Ability/Permission modal was

produced before Tense was overtly realized. We see however where Modal wi was used


140

productively before the productive use of Tense. In addition although all the aspectual

markers examined were attested before Tense, only the Progressive and Prospective markers

were used productively before the productive use of Tense. If children were building the

functional hierarchy from a strict bottom-up approach, we would expect all the root modals

and the aspectual markers to be attested and used productively before the productive use of

Tense. The data therefore does not lend support to the ‘incremental structure building

approach’ to language development. We see however that children acquiring JC never have a

TMA system that contains epistemic modals (the highest element in the clausal hierarchy),

but no aspectual elements (the lowest elements, located closest to the verb).

With regards to the combination of the markers, despite the scarcity in the data, one major

finding was revealed. Children acquiring JC are always target-consistent with respect to the

sequence of utterance. TMA elements which are located in higher clausal layers were never

realized below elements which are lower in the clausal domain. This shows that children from

an early age are aware of the rules governing the cartographic sequencing of the entire TMA

zone. Such finding argues in favor of a ‘full competence’ approach to development.

It appears that the acquisition order may be also influenced by the input as the distributions of

both are skewed in a similar direction. The categories of progressive aspects, prospective

aspects and ability/permission modals are proportionally larger than all other TMA categories.

However the fact that children initially produce more progressive aspectual markers and less

prospective aspectual markers and ability/permission modals than in the input shows that

children do not exactly copy the distribution of the input, and as such the input alone cannot

account for the attested order.


141

In accounting for the relative order attested in the children’s development of the TMA zone,

we propose that all the structure is available from the start of production in line with the full

competence hypothesis. Despite the availability of structure, children nonetheless omit

markers revealing a root infinitival stage in early JC. These omissions are in keeping with the

full competence approach and are based on grammatical options available to the child, in line

with UG. Rizzi (1993/1994) proposed that the option available to the child to ‘strip off’

external clausal layers is responsible for the attestation of root infinitives in early grammars.

Note however, that as soon as this option is no longer available to the child, around 3 years of

age, children start to use the obligatory aspectual markers more productively and root

infinitives decline.

An obvious question is why are some markers more productively used than others, being that

all the structure is available to the child? We propose that, independent of grammatical

options of truncating clausal levels, children are apparently aware that not all markers are

obligatory and hence their low attestation, in line with the input data. But why is the

progressive marker the first to be acquired and used productively? This we argue is due to

semantic reasons. According to Brown & Bellugi (1964) children primarily communicate

about the ‘here-and-now’, and as such initially acquire only the forms they need to do so.

Progressives refer to here-and-now and are used in describing ongoing activities. In keeping

with Boland (2006) operators that are communicatively more relevant and are cognitively less

complex are easier to acquire. The analysis of the input shows where the progressive aspect is

the most dominant TMA element in the communication of young children and as such is most

relevant in their early development, hence acquired first.


142

6.8.3 Conclusions

The analysis of spontaneous speech of children acquiring the TMA system in JC showed that

the obligatory aspectual markers (progressive and prospective) are the first elements to be

produced and productively used. These markers form the lowest level in the TMA zone, and,

importantly, children do not go through a stage where the highest elements in the clausal

hierarchy are produced before the lowest elements. The bottom-up structure building

approach however cannot account for the fine-grained development of the TMA zone in JC as

children do not necessarily acquire all aspectual markers before root modals and tense, nor

will they acquire all root modals before epistemic modals. The empirical finding that children

never produce target-inconsistent TMA combinations provides evidence that children

acquiring JC are aware of the cartographic structure of the entire TMA domain from an early

age.

6.8.4 Future Research

Based on a limitation of the current corpus, in that the input data was not coded, we were

unable to provide a very detailed discussion regarding TMA in child speech and its

correlation in child directed speech. As a result many questions remain open. One such

question is whether there are omissions of TMA markers in the spontaneous speech of adults.

If there are no omissions this would reinforce the view that structures with omitted aspectual

markers are indeed RI. Additionally, statistical analysis of the correlation between the input

and the child speech is a way forward to better understanding the role of input in language

acquisition.

Chapter 7: Null Subject Phenomenon

143

CHAPTER 7

NULL SUBJECT PHENOMENON22

7.1 INTRODUCTION

The goal of this chapter is to provide a systematic analysis of early subject omission in JC, a

non-null subject language23

. Starting from the observation that early subject drop is robustly

attested for several months (as indicated in the previous chapter), the examination of the

omissions reveals a very restricted environment of occurrence. Early subject omission is

essentially confined to the clause initial position, being virtually absent from instances of wh-

preposing, as has been observed for other languages.

The chapter is organized as follow: Section 7.2 discusses the object/subject asymmetry,

showing that while subject drop is robustly attested, object drop is minimal. Section 7.3

describes the general pattern of the null subject phenomenon in the production of the children

in the corpus and situates this in the literature. In section 7.4, detailed evidence that early

subject omission in JC is highly restricted is provided: early subject drop is only possible in

clause-initial position, and virtually excluded after an overt wh-element. This therefore

provides empirical support for the claim that early null subjects are a case of the “Privilege of

the Root”, and for the Truncation Hypothesis (Rizzi 1992; 1993/94; 2006). Section 7.5

examines the phenomenon of null wh-constituent questions and yes/no questions, and shows

that subject drop is possible after a null +Q element, thereby requiring a revision of the

approach to the Privilege of the Root and Truncation. Section 7.6 explores such a revision

22 This chapter includes work submitted for publication co-authored with Luigi Rizzi, Stephanie Durrleman and Ur Shlonsky. 23

Although JC is usually described as a non-null subject language, target-consistent subject drop is attested in some restricted context, for example if the subject was pre-mentioned in the discourse.


144

along the lines of the spell-out mechanism of Phase Theory, rather than on structure building

operations. Section 7.7 summarizes and concludes the chapter.

It should be noted that only verbal utterances have been taken into account for the present

analysis, and only target-inconsistent omissions are included in the count for null elements.

As such imperative clauses were excluded from the analysis of null subjects. Importantly

however, imperative clauses constitute an important percentage of the children’s production

demonstrating target-inconsistent object omissions and as such they were taken into account

in the discussion of null objects in Section 7.2.

7.2 OBJECT / SUBJECT ASYMMETRY

In comparison to the extensive research conducted on subject drop, the phenomenon of object

omission has remained less documented. Some researchers found that they almost did not

exist in some languages (Hyams 1986, Hyams & Wexler 1993, Hamann 1996). Hyams (1986)

shows that while subject omission is frequent (in English), object omission is not. She

proposed that object omissions are more likely to be related to performance errors than to be a

product of a regular grammatical process. Hyams & Wexler (1993) compared proportions of

subject drop with object drop for child English and found a huge discrepancy: subjects were

dropped around 50% of the time, while object drops accounted for merely 10%. Other studies

on object drop in Child English reveals even lower percentages (Wang, Lillo-Martin, Best &

Levit, 1992; Bloom, 1990). Similarly, Hamann (1996) found that null objects nearly do not

exist in German. Additionally, there are large differences reported in the rate of subject and

object omission even for null-object languages, with subject omission still being more robust

than object omission. For Chinese, Wang et al. (1992) reported 56% subject omission


145

compared to 23% object omission; For Korean, Kim (2000) found 77% subject omission

compared to 51% object omission; and for K’iche’, Pye (1992) found 92% and 67% subject

and object omissions respectively.

The main goal of this section then is to assess the extent to which object drop is attested in

early JC and to ascertain whether it is a separate phenomenon than that of subject drop.

Although JC is not usually described as a null object language, target-consistent object drop is

attested in a variety of context. Given the possibility that omissions are highly context-

dependent, the context in which the null elements occurred is used as the main criterion for

establishing target inconsistency. To calculate the target-inconsistent null elements, all

utterances for which the object and/or subject is obligatory based on the discourse are counted

and assessed individually for acceptability. Note that this is the only section where

imperatives are considered. They are included as an important percentage of the children’s

production demonstrating target-inconsistent object omissions are found in these clauses. To

distinguish imperatives from non-imperatives, again the context was the main criterion: for

example imperatives are in the present tense, used for commands and with optional subject.

Moreover, native speakers’ intuitions are also employed in establishing target inconsistency

among the clauses.

For the analysis below, we counted all verbal utterances in the entire corpus. We then counted

all the cases where obligatory subjects and objects were dropped. The findings reveal that

total target-inconsistent null objects accounted for only 2.88% of the entire data while null

subjects accounted for 27.18%. As displayed in Table 7, the subject/object asymmetry is

clearly evident and homogeneous in all informants, despite minor individual variation.


146

INFORMANTS TOTAL VERBAL

UTTERANCE NON-TARGET

NULL OBJ % NULL OBJ

NON-TARGET NULL SUBJ

%NULL SUBJ

ALA 5485 131 2.39% 1198 21.84%

COL 3660 92 2.51% 1038 28.36%

KEM 3896 121 3.11% 1419 36.42%

RJ 3522 101 2.87% 1011 28.71%

SHU 2613 43 1.65% 426 16.30%

TYA 999 93 9.31% 392 39.24%

TOTAL 20175 581 2.88% 5484 27.18%

Table 1: Total Target-inconsistent Omissions in Data Set

The question that arises then is why are the figures drastically lower than those reported for

English? We assume this may be due to differences in the procedures for the calculation. To

rectify this, thereby presenting a tighter analysis of our data with that reported for English, we

conducted a comparative analysis replicating Bloom’s (1990) study. Bloom tested the

prediction that subject omission is ‘selective’ in contrast to object omission by comparing

children’s rates of subject omissions with object omission. He counted utterances which

contained verbs that required obligatory objects in the natural productions of children

acquiring English, for a set of verbs (listed in Table 2) and compared it with omitted subjects

from a set of verbs which requires the subject (listed in Tables 3 and 4). Table 3 includes

verbs that denotes cognitive states or involuntary acts, which he referred to as non-imperative

verbs; and Table 4 includes past tense verbs, since these verbs cannot be used in an imperative

form. Sentences from these lists, not containing a subject, would be true examples of target-

inconsistent subject omission. He found a significant discrepancy: 55% of subjects were

omitted, while only 9% of obligatory objects were.


147

Table 2: Bloom’s 1990 Obligatory object verbs list

Table 3: Bloom’s 1990 Non-imperative Verbs List

Table 4: Bloom’s 1990 Past tense verbs list

A similar analysis on the present corpus was conducted for the same period as per Bloom’s

informants, i.e. informants up to 31 months of age. We selected all the verbs from our data

that corresponds to Bloom’s Tables 2, 3 and 4 above, as presented in Tables 5, 6 and 7 below

respectively. Note that some of the verbs were not found in the present corpus, and as such

were ignored. For the null subject analysis, verbs that fell into both classes were counted as

Past tense verbs. Some of the verbs in Bloom’s lists are ungrammatical in Adult English, due

to overextension of the past tense morpheme. Like in Bloom’s analysis, to calculate null

Bought Drinked Ironed Miss Saved Throwed

Broke Fix Like Need Saw Took

Brought Folded Love Pulled See Want

Caught Found Loves Rode Sharpened Wants

Covered Gave Made Said Thought Washed

Care Grow Live Need

Cry Know Lives See

Fall Laugh Love Sneeze

Falls Laughs Loves Want

Forget Like Miss Wants

Ate Closed Falled Goed Pulled Sharpened Tored

Bit Cooled Fell Ironed Rode Spilled Tripped

Bought Covered Fixed Left Said Stepped Turned

Broke Cried Folded Lost Sat Stopped Washed

Brought Drinked Forgot Made Saved Thought Went

Came Dropped Found Melted Saw Throwed Wrote

Caught Dropt Gave Pee-peed Sent Took


148

elements, imperatives, questions, statement with a negative element, statements where the

verb is part of an embedded clause, and repetitions were not included. To calculate null

subjects and objects we examined verbs that require an obligatory subject or object based on

the discourse. The same criteria for determining target inconsistency in JC as already outlined

were employed. The findings reveal a similarly striking discrepancy: null subjects accounted

for 50.78% of the selected data set for the verbs listed in Tables 6 and 7, while null objects

accounted for only 6.53% of the verbs in Table 5.

Table 5: Obligatory Objects Verb List

Verbs # of Utterances Null Objects

bai 'bought' 49 10

brok 'broke' 26 1

bring 'brought' 0 0

chuo 'threw' 4 2

fiks 'fix' 15 1

fain 'found' 33 7

gi/v 'gave' 17 2

jringk 'drank' 5 1

kech 'caught' 18 1

kova 'covered' 1 0

laik 'like' 14 1

lov 'love/s' 13 0

mek 'made' 17 0

niid 'need' 1 0

pul 'pulled' 4 1

raid 'rode' 6 0

se 'said' 18 0

si 'see/saw' 67 2

tek 'took' 51 6

waa/hn 'want/s 228 4

wash 'washed' 10 0

TOTAL 597 39

6.53%% null object


149

Table 6: Nonimperative Verb List

Verbs # of Utterances Null Subjects

laaf 'laugh/s' 1 0

laik 'like' 14 7

liv 'live/s' 5 1

lov 'love/s' 13 2

niid 'need' 1 0

nuo 'know' 4 4

waa/hn 'want/s' 228 120

TOTAL 266 134

50.38%% null subject


150

Table 7: Past tense Verb List

Verbs # of Utterances Null Subjects

baal 'cried' 1 0

bait 'bit' 34 16

bai ' bought' 50 25

brok 'broke' 38 15

chuo 'threw' 4 2

dash 'spilled' 27 21

faal 'fell' 4 2

fiks 'fixed' 16 4

fain 'found' 33 23

gi 'gave' 17 1

go 'went' 33 14

it/iit 'ate' 44 32

jringk 'drank' 5 3

jrap 'dropped' 270 161

kom 'came' 44 15

kech 'caught' 20 12

kuul 'cooled' 2 0

kova 'covered' 1 1

krai 'cried' 11 6

kyari 'brought' 3 0

lef 'left' 4 2

laas 'lost' 7 3

mek 'made' 17 4

nyam 'ate' 1 0

piipi 'pee-peed' 7 5

pul 'pulled' 4 3

raid 'rode' 6 1

rait 'wrote' 16 9

se 'said' 18 5

sidong 'sat' 1 1

si 'saw' 7 4

shot 'closed' 2 0

tep 'stepped' 1 0

(s)tap 'stopped' 14 4

tek 'took' 55 27

tier 'tore' 8 2

torn 'turned' 11 5

wash 'washed' 14 7

TOTAL 850 435

51.18%% null subject


151

As the fundamental results of the asymmetry between object drop and subject drop remains,

the discrepancy between the figures of two analyses is plausibly due to differences in

calculation procedures. As seen in the latter, the data covered only up to 31 months of age, the

period where null subjects are still robustly attested. On the other hand, the previous analysis

examined the subject/object asymmetry throughout the entire corpus, grouping togething the

periods where null subjects are robust with periods when they are sparsely attested.

Additionally, restricting the verbs for inclusion in the analyses will yield higher numbers of

null objects and subjects as compared to an analysis where verb type and utterance type was

not controlled.

Typical examples of null objects in the corpus included the following:

1) Ø fiks Ø. (ALA 1;11)

Ø fix Ø

“(I am) fixing (it).”

2) Momi mi dalli waahn Ø. (ALA, 2;02)

Mommy my doll want Ø

“Mommy, my doll wants (breast).

3) Momi bied Ø. (COL 1;11)

mommy bathe Ø

“Mommy bathed (me).”

4) A kyahn pin Ø. (COL 2,05)

1SG cannot spin Ø


152

“I cannot spin (it).”

5) Momi biit Ø. (SHU 2;02)

mommy beat Ø

“Mommy beats (him).”

6) Mi waahn put iin Ø. (SHU 2,02)

1SG want put in Ø

“I want to put (it) in.”

7) Ø waa tek out Ø. (RJU 2;07)

Ø want take out Ø

“(I) want to take (it) out.”

8) Pul out Ø! (RJU 1;11)

Pull out Ø

“Pull (it) out!”

9) A mii av Ø. (KEM 2;10)

FOC 1SG have Ø

“I am the one who have (it).”

10) Mi a_go raid mi Ø. (KEM 2;09)

1SG PROS ride my Ø

“I am going to ride my (bike).”


153

11) Ø put aan Ø. (TYA 2;08)

Ø put on Ø

“(I) am going to put (it) on.”

12) Im naa kyaar Ø. (TYA 2;07)

3SG NEG~PROG carry Ø

“She is not taking (anything).”

The above examination provides evidence in favor of the hypothesis that the phenomenon of

subject drop is quite separate and apart from that of object drop, and as such requires a

separate analysis. We will now turn to the discussion of the null subject phenomenon.

7.3. EARLY SUBJECT DROP IN JC

In this section, we examine the phenomenon of null subjects in the acquisition of JC. The

phenomenon of early subject drop in creole languages is an understudied topic, and aside

from Syea’s (1993) and Adone’s (1994) work on Mauritian Creole, and more recently Costa

and Pratas’ (2012) work on Capeverdean, there is no comprehensive discussion of the

occurrence of null subjects in these contexts. However both Mauritian and Capeverdean

deviate from the classical creole pattern by themselves allowing null subjects in certain

contexts. This is not the case in the vast majority of Creole grammars. Atlantic creoles, such

as Guadeloupean, Martinican, Belizean, Berbice Dutch, Guyanese, Saramaccan, Sranan,

Haitian Creole, to name a few, show a very strong tendency to have obligatory subject

pronouns (Haspelmath, et.al 2013). Likewise, JC requires the overt expression of subjects.


154

The general conclusion is then that ‘more cross-linguistic data is required to provide more

insights into this domain, together with the study of creole acquisition’ (Adone, 1994:144).

The pattern observed in other languages is that children go through a stage in their language

development where they omit subjects. Subject omission occurs both in the acquisition of null

subject languages and non-null subject languages, even though at different rates and with

different structural characteristics. This stage typically occurs for several months but usually

tend to gradually disappear around the end of the third year of life (Rasetti 2003). For

instance, Hamann, Rizzi & Frauenfelder (1996) examined early omission of subjects in

French and found that null subjects fall under the bar of 20% only around 2;10. This

phenomenon is also attested in the early JC linguistic productions, where children tend to omit

the requirement of overtly expressing the subject, thereby producing utterances with root null

subjects as exemplified in (13) – (24). All the following sentences would be ungrammatical in

adult Jamaican:

13) Ø iit ais-kriim. (COL 1;11)

(1SG) eat ice-cream

“(I) ate ice-cream.”

14) Ø jraiv i tu. (RJU 2;01)

(2SG) drive 3SG too

“(You)drove it also.”

15) Ø fit mi fingga. (SHU 2;03)

(3SG) fit 1SG finger

“(It) fits my finger.”


155

16) Ø jraiv van. (KEM 2;04)

(1PL) drive van

“(We) drove the van.”

17) Ø taak tu dem aredi. (ALA 2;09)

(2PL) talk to them already

“(You) spoke to them already.”

18) Ø iit fuud don. (TYA 2;08)

(3PL) eat food COMP24

“(They) finished eating the food.”

Figures 1 to 6 detail the individual production of target-inconsistent null subjects in our

corpus. This includes both declaratives and interrogatives for which the subject is omitted. As

previously specified, imperatives are excluded from this analysis. On the x-axis we present

the percentage production of null subjects and on the y-axis the informants’ age.

24

COMP is used to refer to the completive aspect marker.


156

Figure 1: COL’s null subjects Figure 2: ALA’s null subjects

Figure 3: RJU’s null subjects Figure 4: TYA’s null subjects

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1;8,

17

1;9,

28

1;11

,12

2;0,

28

2;2,

16

2;3,

30

2;5,

14

2;6,

25

2;8,

6

2;9,

24

2;11

,70%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1;9,

25

1;11

,5

2;0,

20

2;2,

6

2;3,

24

2;5,

7

2;6,

22

2;8,

2

2;9,

14

2;11

,1

3;0,

15

ALA'S NULL SUBJECTS

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1;1

0,4

1;1

1,1

5

2;0

,30

2;2

,16

2;4

,4

2;5,

17

2;7

,5

2;8

,12

2;9

,24

2;1

1,1

1

3;0

,25

RJU'S NULL SUBJECTS

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1;1

1,2

5

2;1

,5

2;2,

20

2;4

,6

2;5

,24

2;7,

7

2;8

,22

2;1

0,2

2;11

,14

3;1

,1

3;2

,15

TYA'S NULL SUBJECTS

COL'S NULL SUBJECTS


157

Figure 5: KEM’s null subjects Figure 6: SHU’s null subjects

As demonstrated, there is a steady decline in the use of null subjects by all participants, not

surprisingly with some individual variation. All participants, with the exception of SHU,

displayed initial production of null subjects over a high of 60%, which all gradually fall, in

some cases to a low of under 10%. SHU, the eldest informant studied, as expected, had the

lowest production of null subjects throughout the corpus. Note that at her first recording her

MLU was already at 2.9, as opposed to other informants whose MLU reaches that point at a

much later age. For example TYA did not reach an MLU of 3 until 2 years and 8 months of

age.

The examination of the total production of null subjects reveals a striking correspondence

with results presented in Hamann, Rizzi and Frauenfelder’s (1996) study. As demonstrated in

Figure (7), at 2;08 the production of null subjects drops to 20%, and it is just above 10% at

3.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2;0,

21

2;2,

1

2;3,

16

2;5,

2

2;6,

20

2;8,

3

2;9,

18

2;10

,29

3;0,

10

3;1,

28

3;3,

11

KEM'S NULL SUBJECTS

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

2;1,

23

2;3,

3

2;4,

18

2;6,

4

2;7,

22

2;9,

5

2;10

,20

3;0,

0

3;1,

12

3;2,

30

3;4,

13

SHU'S NULL SUBJECTS


158

Figure 7: Percentage null subjects in early JC

It is our assumption that around 36 months (when null subject production fluctuates

constantly around 10%) the child system converges to the target grammar. At this point null

subjects are less productive. It is possible that these null subjects need not be accounted for

given that they correspond to the adult type of subject omission. It therefore appears to be a

just decision to extend our analysis only up to 35 months when the phenomenon is more

robust in the corpus. Provided in tables (8) – (13), are detailed breakdown of the number of

omissions of obligatory subjects with respect to the total production of verbal utterances for

each participant. Like in Figures (1) to (6), we examine declaratives and interrogatives,

omitting imperative utterances.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

1;8.

0

1;9.

0

1;10

.0

1;11

.0

2;0.

0

2;1.

0

2;2.

0

2;3.

0

2;4.

0

2;5.

0

2;6.

0

2;7.

0

2;8.

0

2;9.

0

2;10

.0

2;11

.0

3;0.

0

3;1.

0

3;2.

0

3;3.

0

3;4.

0

AGE (Y;M)

% N

ULL

SU

BJE

CTS


159

AGE (Y;M,D) UTTERANCES

NULL SUBJ

%NULL SUBJ

1;8,17 24 15 62.50

1;8,27 27 17 62.96

1;9,17 34 21 61.76

1;9,28 53 32 60.38

1;10,8 36 18 50.00

1;11,1 47 29 61.70

1;11,12 56 39 69.64

1;11,28 60 36 60.00

2;0,12 47 26 55.32

2;0,28 70 39 55.71

2;1,14 75 45 60.00

2;2,0 46 28 60.87

2;2,16 155 85 54.84

2;3,1 98 44 44.90

2;3,16 103 65 63.11

2;3,30 113 54 47.79

2;4,15 92 34 36.96

2;5,0 100 25 25.00

2;5,14 71 23 32.39

2;5,27 127 33 25.98

2;6,10 72 21 29.17

2;6,25 207 26 12.56

2;7,8 156 40 25.64

2;7,22 113 31 27.43

2;8,6 178 51 28.65

2;8,20 179 24 13.41

2;9,11 68 23 33.82

2;9,24 87 23 26.44

2;10,10 116 24 20.69

2;10,21 122 30 24.59

2;11,7 130 19 14.62


NULL SUBJ

%NULL SUBJ

1;9,25 47 31 65.96

1;10,4 33 27 81.82

1;10,25 101 86 85.15

1;11,5 88 72 81.82

1;11,16 82 62 75.61

2;0,9 97 60 61.86

2;0,20 62 53 85.48

2;1,5 87 76 87.36

2;1,20 85 53 62.35

2;2,6 116 69 59.48

2;2,22 187 82 43.85

2;3,8 44 11 25.00

2;3,24 130 61 46.92

2;4,9 121 69 57.02

2;4,24 151 54 35.76

2;5,7 111 25 22.52

2;5,23 249 58 23.29

2;6,12 80 13 16.25

2;6,22 74 9 12.16

2;7,5 178 15 8.43

2;7,18 172 16 9.30

2;8,2 278 32 11.51

2;8,16 188 7 3.72

2;9,0 160 13 8.13

2;9,14 159 19 11.95

2;9,28 160 12 7.50

2;10,25 173 13 7.51

2;11,1 159 28 17.61

2;11,18 155 25 16.13

TABLE 8: COL’S NULL SUBJECTS TABLE 9: ALA’S NULL SUBJECTS


160


NULL SUBJ

%NULL SUBJ

1;10,4 8 7 87.50

1;10,14 14 13 92.86

1;11,4 68 54 79.41

1;11,15 17 8 47.06

1;11,26 20 16 80.00

2;0,19 44 30 68.18

2;0,30 70 50 71.43

2;1,15 95 46 48.42

2;2,0 112 72 64.29

2;2,16 64 26 40.63

2;3,1 169 67 39.64

2;3,18 157 57 36.31

2;4,4 89 29 32.58

2;4,19 99 38 38.38

2;5,3 91 41 45.05

2;5,17 112 45 40.18

2;6,2 116 25 21.55

2;6,18 126 33 26.19

2;7,5 195 32 16.41

2;7,15 172 29 16.86

2;7,28 112 25 22.32

2;8,12 183 47 25.68

2;8,26 117 23 19.66

2;9,10 142 51 35.92

2;9,24 110 26 23.64

2;10,7 109 15 13.76

2;10,29 74 24 32.43

2;11,11 84 24 28.57

TABLE 10: RJU’S NULL SUBJECTS

AGE (Y;M,D)

UTTERANCES NULL SUBJ

%NULL SUBJ

1;11,25 11 9 81.82

2;0,4 25 23 92.00

2;0,25 9 6 66.67

2;1,5 36 33 91.67

2;1,16 18 16 88.89

2;2,9 13 8 61.54

2;2,20 28 24 85.71

2;3,5 22 20 90.91

2;3,20 14 12 85.71

2;4,6 20 15 75.00

2;4,22 14 11 78.57

2;5,8 7 7 100.00

2;5,24 19 15 78.95

2;6,9 15 6 40.00

2;6,24 50 26 52.00

2;7,7 2 1 50.00

2;7,23 43 21 48.84

2;8,8 74 34 45.95

2;8,22 29 9 31.03

2;9,5 64 20 31.25

2;9,18 13 7 53.85

2;10,2 48 11 22.92

2;10,16 44 11 25.00

2;11,0 69 12 17.39

2;11,14 99 12 12.12

TABLE 11: TYA’S NULL SUBJECTS


161


NULL SUBJ

%NULL SUBJ

2;0,21 83 64 77.11

2;1,0 88 71 80.68

2;1,21 87 82 94.25

2;2,1 87 82 94.25

2;2,12 132 124 93.94

2;3,5 111 96 86.49

2;3,16 92 77 83.70

2;4,1 129 81 62.79

2;4,16 74 58 78.38

2;5,2 83 61 73.49

2;5,18 121 100 82.64

2;6,4 90 41 45.56

2;6,20 125 56 44.80

2;7,5 192 72 37.50

2;7,20 151 58 38.41

2;8,3 138 28 20.29

2;8,19 142 10 7.04

2;9,4 195 24 12.31

2;9,18 141 23 16.31

2;10,1 76 13 17.11

2;10,14 125 31 24.80

2;10,29 221 44 19.91

2;11,12 141 33 23.40

TABLE 12: KEM’S NULL SUBJECTS

AGE (Y;M,D) UTTERANCES NULL SUBJ

%NULL SUBJ

2;1,23 77 23 29.87

2;2,2 99 16 16.16

2;2,11 46 11 23.91

2;3,3 91 52 57.14

2;3,14 119 52 43.70

2;4,7 30 3 10.00

2;4,18 28 2 7.14

2;5,3 46 9 19.57

2;5,18 137 30 21.90

2;6,4 132 29 21.97

2;6,20 110 13 11.82

2;7,16 100 25 25.00

2;7,22 99 21 21.21

2;8,7 97 8 8.25

2;8,22 64 6 9.38

2;9,5 110 10 9.09

2;9,21 142 15 10.56

2;10,6 74 16 21.62

2;10,20 26 1 3.85

2;11,3 122 13 10.66

2;11,16 64 9 14.06

TABLE 13: SHU’S NULL SUBJECTS

Several proposals have been put forward to account for this robust attestation of early subject

drop across languages which is not permitted by their target grammar, ever since Hyams’

(1986) seminal work on the topic. Hyams proposed the Pro hypothesis, suggesting that the

pro-drop parameter is initially set to the positive value and later resets in accordance with

evidence from the target language. This account proved to be problematic as it assumed that

child null subjects are similar to adult null subject languages. The prediction is not borne out

as evidence reveals that the distribution of null subjects in child languages is very restricted in


162

comparison with adult pro-drop languages (Valian 1990, Rizzi 2002). The same problem

arises for the version of the parametric approach adopted by Yang (2002) to capture the early

null subject phenomenon. Hyams and Wexler (1993) used topic-drop to account for the null

subject phenomenon in child language. They proposed that missing subjects are due to the

same processes allowing topicalized constituents to be dropped in a topic-drop language

(Hyams and Wexler, 1993). Again, there are drawbacks to accepting a purely topic-drop

account, such as the low rate of topicalized object drops compared to subject drop.

Based on empirical observations due to Valian (1990), Rizzi (1992) conjectured that early

subject drop is restricted to the clause-initial position, an instance of a more general property

making it possible to leave the specifier of the root node unpronounced; this has been called

the “Privilege of the Root”. He introduced the Truncation approach as a way to make

structurally possible this particular instance of Privilege of the Root. If children have the

option of truncating the CP layer in main declaratives, the subject position can become the

specifier of the root node, a position which has the privilege of remaining unpronounced and

accessible to discourse identification, thus escaping normal sentence-internal identification

requirements for empty elements. Indeed the requirement that null elements have a clause

internal identifier only applies if it is virtually satisfiable. Given that the specifier of the root is

not c-commanded by any category, it has no potential structural identifier and is thus

exempted from clause-internal identification. Crucially, the Truncation hypothesis proposes

that while an immature system initially does not systematically project to the CP layer, if and

when it does, all the intermediate projections must be present as well. It therefore follows

from the Truncation approach that early null subjects are virtually absent in interrogatives

involving wh-preposing, since these obligatorily contain a CP and therefore the subject

position would no longer qualify as the specifier of the root. For additional arguments that


163

null subjects in child language is a grammatical phenomenon, one may consult Lillo-Martin

(1986), Lebeaux (1987), Jaeggli & Hyams (1988), Pierce (1992), Sano & Hyams (1994),

Rizzi (1994, 2000, 2002, 2006) Boomberg & Wexler (1995), Roper & Rohrbacher (1994,

2000) and Wexler (2014) among others.

On the other hand, rather than relating developmental stages to theories of Universal

Grammar, other approaches accounting for null subjects are performance-based, claiming that

children’s grammar is essentially like that of adults, however children’s productions are not

always faithful to this internalized grammar because of processing constraints. In support of

this view of a discrepancy between performance and competence, Bloom (1990) argues that

children know more about their target grammar (its rules and structures) than they themselves

are able to produce. According to this perspective, omissions in production may be due to the

child’s capacities being overloaded by the length and structural complexity of utterances (L.

Bloom 1970), which leads to an extra-grammatical simplification through the omission of the

subject; contextual information may lead to recovering the meaning of the omitted constituent

(Greenfield & Smith 1976). Important evidence against a purely performance-based approach

has been provided by Orfitelli and Hyams (2008), who show that children in the subject drop

stage also accept subjectless sentences in English. This suggests that the null subject stage is

not merely a production phenomenon, but also a comprehension one, which requires a

grammar-based approach which can predict parallel production and comprehension results.

In this chapter, we argue in favor of a system that analyses early subject drop as a

grammatical phenomenon, in line with the tradition initiated by Hyams (1986). According to

this view, the strategies of ellipsis applied by the child are UG-consistent operations, or

parametric values, also present in some adult languages. The causal factor leading the child to


164

temporarily entertain such values may well be linked to performance limitations, as such non-

target-consistent values typically place minimal burden on the child’s production system

(Rizzi 2005a); nevertheless the crucial point in this approach is that the child uses a fully

legitimate grammatical option, made available by UG.

The JC data considered above provide clear evidence for a null subject phase in JC. It is

expected however that these null subjects should be limited to occur in the specifier of the

root, and as such will be virtually absent from wh-questions involving wh-preposing of a non-

subject element. This will be examined in the next section.

7.4. SUBJECT OMISSION AS A MANIFESTATION OF THE PRIVILEGE OF

THE ROOT: OVERT WH-QUESTIONS.

Recall that Hyams (1986) put forth the hypothesis that early subject drop results from a mis-

setting of the Null Subject Parameter. Rizzi (1992) however, argued that the early subject

omission of a non-null subject language (non-NSL) does not correspond to the omission of

subjects of true null-subject languages, based on Valian’s (1990) observation that subject

omission is quite robust in early English declaratives, but it is virtually absent in post wh

environments. This finding gave rise to the following conjecture:

19) Early subject drop in the acquisition of a non-null subject language is only possible

in the Specifier of the root.

If the sentence starts with a wh-element in the C system, the subject is not the specifier of the

root, and cannot be dropped. This is true also for other languages: For example French

(Crisma, 1992; Levow 1995), Dutch (Haegeman 1995, 1996a), and German (Clahsen,

Kursawe and Penke, 1996). These findings provided evidence that early subject drop in non-


165

NSL is a separate phenomenon from the positive fixation of the null subject parameter as their

structural environments are different. Null subject languages allow null subjects in initial and

non-initial (post-wh and embedded) environments, as shown in example (20) from Italian,

while early subject drops in non-NSL are restricted:

20) Dove Ø va

Where Ø goes?

JC, being a non-NSL, is expected to be in keeping with conjecture (19) above. Indeed, during

the period of high target-inconsistent null subject production (up to 35 months), the omission

of subjects is highly restricted.

Subjects are only dropped in initial positions and hardly ever in contexts following a wh-

element. So, in post-wh environments we typically find overt subjects, as in the following

cases:

21) Wa im a ron fa? (ALA 2;06)

why 3SG PROG run for

“Why is he running?”

22) We Matyu gaan? (ALA 2;07)

where Matthew gone

“Where has Matthew gone?”

23) We momi kom fram? (COL 2;02)

where mommy come from

“Where mommy came from?”


166

24) Wa granmaa a du? (COL 2 ;05)

what grandma PROG do

“What is grandma doing?”

25) We yaa luk ova de so? (KEM 2;09)

why 2SG~PROG look over there so

“Why are you looking over there?”

26) We i kyat a du, mm? (KEM 2;08)

what DET cat PROG do COM

“What is the cat doing?”

27) Uu shi a taak ? (RJU 2;07)

who 3SG PROG taak

“Who is she talking to?”

28) We dem a go? (RJU 2;07)

where 3PL PROG go

“Where are they going?”

29) Ou yo lak i? (SHU 2;09)

how 2SG lock 3SG

“How do you lock it?”


167

30) We yo tikl mi fa? (SHU 2;05)

why 2SG tickle 1SG for

“Why did you tickle me?”

31) We Moiisha a kaal mi fa? (TYA 2;09)

what Moesha PROG call 1SG for

“What is Moesha calling me for?”

32) We yaa go? (TYA 2;06)

where 2SG~PROG go

“Where are you going?”

Let us now examine the acquisition of interrogatives in JC and its interaction with early null

subjects. We will center our analysis around the period where the production of null subjects

is still robust in the corpus, i.e. up to 2;11, and present evidence in support of the argument

that early null subjects in JC are a manifestation of the Privilege of the Root.

7.4.1 Overt wh-phrase and Null Subjects

Table 14 shows the production of overt wh-elements by children up to the age of 2;11. In

order to provide an accurate representation of the subject omissions in the data, 162 utterances

where the wh-element itself is the subject are not presented. During this period where null

subjects are still robustly attested in the corpus, of 900 overt non-subject wh-questions, only

10 cases were noted where the subject was dropped in post-wh position. This represents

1.11% of overt wh-constituent questions for which the wh-element is not the subject. This


168

finding is in keeping with the literature as it has been reported that such sentences normally

occur between just under 1 percent and 5 percent for other languages: French - 0.9% null

subjects in wh-questions compared to 40.5% in declaratives (Crisma 1992), Dutch - 2% null

subjects in wh-questions compared to 23.5% in declaratives (Haegeman 1995), English -

1.6% null subjects in wh-questions (Valian 1991), German - 4% null subjects in wh-questions

(Clahsen, Kursawe and Penke 1995), Swedish - 5.2% null subjects in wh-questions (Platzack

and Josefsson 2000). The minimal attestation of subject drop following a wh-element shows

that in JC, like the above mentioned non-NSLs, subject drop is restricted to the initial

position.


169

AGE COL WH ALA WH RJU WH TYA WH KEM WH SHU WH TOTAL WH

(Y;M) +Sub -Sub +Sub -Sub +Sub -Sub +Sub -Sub +Sub -Sub +Sub -Sub +Sub -Sub

1;8.0 0 0

0 0

1;8.5 0 0

0 0

1;9.0 0 0

0 0

1;9.5 0 0 1 0

1 0

1;10.0 0 0 2 0 0 0

2 0

1;10.5 2 0 1 0 0 0

3 0

1;11.0 2 0 0 0 0 0

2 0

1;11.5 0 0 8 0 2 0 0 0

10 0

2;0.0 0 0 12 0 0 0 0 0

12 0

2;0.5 1 0 4 0 2 0 0 0 0 0

7 0

2;1.0 0 0 6 0 0 0 0 0 0 0

6 0

2;1.5 0 0 1 0 1 0 0 0 0 0 2 0 4 0

2;2.0 2 0 9 0 0 0 1 0 0 0 8 0 20 0

2;2.5 3 0 7 0 1 1 0 0 0 0 3 1 14 2

2;3.0 0 0 9 0 3 0 0 0 0 0 3 0 15 0

2;3.5 2 0 14 1 4 0 1 0 0 0 4 0 25 1

2;4.0 1 0 4 0 7 0 0 0 1 0 4 0 17 0

2;4.5 0 0 10 4 5 0 0 0 0 0 4 0 19 4

2;5.0 1 0 3 0 6 0 0 0 0 0 9 0 19 0

2;5.5 8 1 20 0 0 0 0 0 0 0 17 0 45 1

2;6.0 1 0 5 0 6 0 0 0 1 0 15 0 28 0

2;6.5 17 0 4 0 17 0 4 0 1 0 15 1 58 1

2;7.0 2 0 13 0 47 1 0 0 19 0 13 0 94 1

2;7.5 5 0 18 0 25 0 0 0 11 0 13 0 72 0

2;8.0 13 0 25 0 1 0 3 0 1 0 13 0 56 0

2;8.5 3 0 15 0 11 0 3 0 3 0 9 0 44 0

2;9.0 3 0 14 0 18 0 5 0 17 0 29 0 86 0

2;9.5 5 0 15 0 12 0 1 0 6 0 28 0 67 0

2;10.0 1 0 14 0 10 0 2 0 3 0 11 0 41 0

2;10.5 2 0 10 0 8 0 1 0 4 0 2 0 27 0

2;11.0 8 0 7 0 2 0 2 0 18 0 33 0 70 0

2;11.5 7 0 4 0 5 0 6 0 4 0 26 0

TOTAL 82 1 258 5 192 2 28 0 91 0 239 2 890 10

Table 14: Production of overt wh and null subjects25

The 10 exceptional cases of subject drop following a wh-clause26

are presented in examples

(33) to (42) below:

25

In this table and throughout the chapter, +sub means “overt subject”, and –sub means “null subject”


170

33) Wai Ø jrap i? (ALA 2;04)

why Ø drop it

“Why did you drop it?”

34) Wier Ø goin Kiisha? (ALA 2;03)

where Ø going Kiisha

“Where are you going Keisha?”

35) We Ø jraiv i fa? (ALA 2;04)

why Ø drive it for

“Why did you drive it?”

26There is one utterance in the corpus with an in situ wh-element and a null subject. In order not to detract from the discussion of null subjects following a wh-element, we present it here: (i) Ø ina i ous fi wa? (KEM 2:09) Ø into DET house for what “Why is she in the house?” Do note that the option for the wh-element to remain in-situ is rarely attested in JC most robustly as an echo-question (see Durrleman-Tame 2008), as in e.g. ii, and as such children rarely explore this option. Only 5 of these utterances were noted in the corpus. (ii) INV: yo afi go aks granmaa fi mek som fi yo leeta. “You have to ask grandma to make some for you later.” CHI: aks uu? (ALA 2;04) Ask who “Ask who?” Additionally we have seen 10 utterances where target-consistent in-situ-wh is permitted in ‘discourse-bound’ restricted contexts, as exemplified in (iii). (iii) CHI: we i bi? “What is it?” INV: rekaada. “recorder” CHI: rekaada? “recorder?” INV: ye. “yes” CHI: fi rekaad uu? (SHU 2;05) INF record who “To record who?”


171

36) Ou Ø nuo? (ALA 2;04)

how Ø know

“How do you know?”

37) We Ø goin? (ALA 2;04)

where Ø going


38) Wa Ø du? (COL 2;05)

what Ø do?

“What is he doing?”

39) We Ø gaan in? (RJU 2;02)

where Ø gone in

“Where did he go in?”

40) We Ø aa go? (RJU 2;07)

where Ø PROG go

“Where are they going?”

41) Uu Ø bi? (SHU 2;02)

who Ø be

“Who is he?”

42) We Ø du? (SHU 2;06)


172

what Ø do

“What did you do?”

We now present a comparative analysis of the production of null subjects in declaratives

during the same period.



1;8.0 9 15

9 15

1;8.5 10 16

10 16

1;9.0 12 20

12 20

1;9.5 21 31 12 26

33 57

1;10.0 18 18 4 27 1 7

23 52

1;10.5 15 29 10 71 1 13

26 113

1;11.0 15 39 5 61 13 53

33 153

1;11.5 24 35 8 56 6 7 2 9

40 107

2;0.0 18 25 20 48 4 16 2 21

44 110

2;0.5 27 36 5 51 12 29 3 5 19 62

66 183

2;1.0 29 41 5 74 19 43 3 33 17 66

73 257

2;1.5 15 28 30 46 45 42 2 15 4 80 50 22 146 233

2;2.0 66 82 32 38 37 64 4 8 5 79 67 14 211 285

2;2.5 50 39 86 54 36 23 4 24 7 110 29 9 212 259

2;3.0 38 61 20 7 96 65 2 20 15 83 31 44 202 280

2;3.5 56 52 45 33 93 53 1 12 12 71 61 44 268 265

2;4.0 54 31 43 58 44 22 5 15 35 73 21 3 202 202

2;4.5 74 24 77 42 55 37 3 11 14 34 18 2 241 150

2;5.0 44 23 75 16 40 38 0 7 21 51 23 8 203 143

2;5.5 75 29 153 52 53 37 4 15 19 88 74 25 378 246

2;6.0 49 21 61 12 77 20 9 6 34 37 74 22 304 118

2;6.5 139 22 53 9 59 28 19 24 65 45 66 9 401 137

2;7.0 97 35 132 13 95 26 1 1 91 66 51 13 467 154

2;7.5 72 29 130 15 103 26 21 21 81 47 53 6 460 144

2;8.0 106 48 191 23 77 22 35 32 105 23 71 5 585 153

2;8.5 138 21 140 3 114 36 16 9 127 8 44 6 579 83

2;9.0 39 23 121 12 67 17 38 20 146 20 62 9 473 101

2;9.5 55 22 106 17 66 47 5 7 108 22 93 11 433 126

2;10.0 84 20 113 11 68 25 35 11 59 12 44 12 403 91

2;10.5 74 24 134 11 85 15 32 11 89 31 23 1 437 93

2;11.0 95 18 117 25 45 23 55 12 145 42 69 12 526 132

2;11.5 113 22 54 22 79 11 96 33 47 8 389 96

TOTAL 1618 957 2041 933 1465 856 380 360 1314 1183 1071 285 7889 4574

Table 15: Production of declaratives and null subjects


173

Table 15 shows a sharp difference in the production of null subjects in declarative clauses

when compared to that attested in overt wh phrases. During the period under examination,

subjects are omitted in 36.7% of declarative utterances compared to 1.1% omission in overt

wh-phrases, where the wh-element is not the subject. The distribution of null subjects appears

to be sensitive to the initial position in the phrase. The presence of the wh-element presents

restrictions on subject drop, even during the phase where the omission of subjects is very

high. As shown in Table 16, during the single-word stage where MLU is <1.49, null subjects

in declarative utterances is at its highest at 83.3%. It gradually falls as MLU increase, and

during the complex construction stage where MLU is 4.5 and above the phenomenon is at

13.9%. Comparing with null subjects in overt wh-constructions, the phenomenon is roughly

concentrated during the first multi-word stage where MLU is 2.5 – 3.49. Nonetheless, it

represents a mere 3.3% during this stage.

MLU (words) Overt Wh +subject

Overt wh null subject

Declaratives + subject

Declaratives null subject

< 1.49 4 0 0.0% 39 195 83.3%

1.5-2.49 46 0 0.0% 471 1566 76.9%

2.5-3.49 263 9 3.3% 2763 1756 38.9%

3.5-4.49 413 1 0.2% 2811 765 21.4%

4.5 > 164 0 0.0% 1805 292 13.9%

TOTAL 890 10 1.1% 7889 4574 36.7%

Table 16: Null subject production with MLU

The comparison between declarative and wh environments is summarized in figure 8.


174

Figure 8: Comparison between Declaratives and Overt Wh-questionS

This striking discrepancy is in line with our expectations and can be accounted for by the

Truncation approach. Children’s early clauses may be truncated at structural layers lower than

CP. When CP is not projected, subjects occupy the highest position of the structure and may

remain null. A declarative utterance may therefore be truncated at the IP level, as shown in

(43), thereby permitting null subjects, a particular case of the Privilege of the Root.

43)

43a) (Mi) iit ais-kriim (COL 1;11)

(1SG) eat ice-cream

“I eat ice-cream.”

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

declaratives + subj null subjdeclaratives

overt wh questions+subj

null subj overt whquestions

CP

IP

iit ais-kriim

(mi)


175

In wh-questions involving wh-fronting, however, the CP must be generated to host the wh-

element. Truncation is therefore not possible; the subject is consequently no longer the Spec

of the root and cannot be dropped. It is during the third year of life that the option to truncate

becomes very restrictive and null subjects disappear.

7.5. NULL SUBJECTS WITH NULL QUESTION-ELEMENTS

If the pattern with overt wh-elements straightforwardly supports the classical Truncation

approach just illustrated, the option of null subjects in questions with a null +Q-element

introduces an important variation on the theme, which requires a revision of the approach. We

will now look at this phenomenon.

7.5.1 Null wh

The literature reports that children frequently drop wh-words when acquiring V2 languages

such as Swedish, Dutch and German, (Santelmann 1995, 1997; Van Kampen 1997; Felix

1980; among others) as exemplified in (44):

44) Ø sa du? (Embla 2;03: Santelmann 1997)

Ø said you

“(What) did you say?”

V2 languages make wh-drop very easily recognizable, precisely because of the V2 constraint.

A question with null wh is an interrogative sentence with the inflected verb in first position

and a clause-internal gap. The interpretation as a wh-question is rendered plausible by the

context. The phenomenon is clearly not restricted to the acquisition of V2 languages, see


176

Yamakoshi (2002) on English, French, and Spanish. In all these cases, wh-drop is not target-

consistent, but there are cases of adult languages manifesting wh-drop in different structural

environments: American Sign Language (Petronio & Lillo-Martin, 1997), Wolof (Torrence,

2012), Norwegian (Svenonius and Kennedy 2006), Bavarian (Bayer 2010). Child wh-drop is

observed during the period where wh-questions with overt wh-words are also produced, so

that overt and null wh-elements alternate in natural production. Yamakoshi’s (2002)

elicitation experiment showed a sizable number of wh-drop in productions by learners of

English, but not in productions by learners of Japanese, suggesting that wh-drop can only

affect XPs in SpecCP, and never wh-elements in situ (as in Japanese). These observations

strongly suggest that at least the core cases of wh-drop are another instance of the Privilege of

the Root.

We have observed that this phenomenon of null wh is also attested in our data. Despite the

requirement of adult JC to overtly express the wh-element in constituent questions, children

omitted the wh-element in 21.2% (286/1348) of their utterances. This is comparable to the

quantitative dimension of the phenomenon in other child languages (For example, Santelmann

2003 observed 19% of wh-drop). (45) - (48) show examples of null wh attested in our corpus:

45) Ø i gorl niem? (COL 2;05)

Ø DEF girl name

“(What is) the girl’s name?”

46) Ø Jada lip-glaas de? (SHU 2;07)

Ø Jada lip-gloss LOC

“(Where is) Jada’s lip-gloss?”


177

47) Ø i likl gorl a go? (ALA 2;07)

Ø DET little girl PROG go

“(Where is) the little girl going?

48) Ø shi a du? (RJU 2;06)

Ø 3SG PROG do

“(What is) she doing?”

Closer examination of the data reveals a strikingly high number of null subjects in these wh-

questions with null wh-elements compared to wh-questions where the wh-element is overtly

realized. We provide a detailed examination of this phenomenon in the next section.

7.5.2 Null Wh-elements and Null Subjects in JC

As revealed by the figures in Table 17 below, when the wh-element is not overtly pronounced

in constituent questions, the subject has the option to be realized or to remain null. This is

attested in 60 out of 265 cases (22.6%) of the null wh-questions where the omitted wh-

element is not the subject. Our data reveals a striking correspondence to that reported by

Clashen et al (1996) for German, where 18% missing subjects were found in null wh-

questions.


178



1;8.0 0 0

0 0

1;8.5 0 0

0 0

1;9.0 0 0

0 0

1;9.5 0 0 3 0

3 0

1;10.0 0 0 0 0 0 0

0 0

1;10.5 0 0 3 0 0 0

3 0

1;11.0 0 0 11 0 1 1

12 1

1;11.5 0 0 4 0 0 0 0 0

4 0

2;0.0 0 0 5 1 0 0 0 0

5 1

2;0.5 0 1 0 0 0 0 0 0 0 0

0 1

2;1.0 0 0 0 0 1 2 0 0 0 0

1 2

2;1.5 0 0 0 0 0 1 0 0 0 0 0 0 0 1

2;2.0 1 1 2 1 0 0 0 0 0 0 2 0 5 2

2;2.5 1 4 4 2 0 0 0 0 1 0 0 0 6 6

2;3.0 0 1 1 0 0 0 0 0 0 0 2 1 3 2

2;3.5 0 0 5 5 1 0 0 0 3 0 0 2 9 7

2;4.0 1 0 0 0 2 0 0 0 7 1 0 0 10 1

2;4.5 1 1 3 1 0 0 0 0 0 1 1 0 5 3

2;5.0 1 0 0 0 2 0 0 0 0 1 0 0 3 1

2;5.5 6 0 3 1 1 1 0 0 0 1 2 1 12 4

2;6.0 1 0 0 0 3 0 0 0 10 0 3 4 17 4

2;6.5 22 2 0 0 2 0 0 1 1 3 6 0 31 6

2;7.0 11 1 1 0 8 2 0 0 6 2 5 1 31 6

2;7.5 2 1 0 0 2 0 1 0 0 1 4 2 9 4

2;8.0 3 0 2 1 0 0 1 0 0 0 0 0 6 1

2;8.5 0 0 1 0 2 0 0 0 0 0 1 0 4 0

2;9.0 1 0 1 0 6 0 1 0 0 0 3 0 12 0

2;9.5 1 0 0 0 7 0 0 0 0 1 0 0 8 1

2;10.0 0 1 0 0 1 0 0 0 0 0 1 2 2 3

2;10.5 3 1 0 0 1 0 0 0 0 0 0 0 4 1

2;11.0 0 0 0 0 0 0 0 0 0 0 0 1 0 1

2;11.5 0 0 0 0 0 1 0 0 0 0 0 1

TOTAL 55 14 49 12 40 7 3 2 28 11 30 14 205 60

Table 17: Production of null wh and null subjects.

Again, we do not include the 21 utterances where the omitted wh-element is the subject.

‘What’, ‘where’ and ‘why’ questions (25, 17 and 16 respectively) are most frequently omitted

in questions that also omit the subject. These are also the most attested question types in the


179

corpus in addition to the ‘who’ questions. We will come to a more comprehensive discussion

of question types in the following chapter.

In order to verify that these utterances are authentic wh-questions, which drop both wh-

element and subject, the context of utterance and the intonation is essential. Provided below

((49) – (54)) are some examples and their context of utterance.

49) INV: Felisha beks wid yo RJ?

“Is Felisha vexed with you RJ?”

CHI: Mmh?

“mmh?”

INV: Felisha veks wid yo, ar a AJ shi veks wid? kaa mi nuo a nommii shi

veks wid.

“Is Felisha vexed with you or is it AJ that she is vexed with? Because

I know it is not me she is vexed with.”

CHI: Ø Ø aa veks wid? (RJU

1;11)

Ø Ø PROG vex with

“Who is she being vexed with?”

50) CHI: Uu dat?

“Who is that?”

INV: waa man.

“a man.”

CHI: Man


180

“man?”

INV: Mmhmm

“yes.”

CHI: Ø Ø kum fram? (ALA 2;02)

Ø Ø come from

“Where does he come from?”

INV: Mi no nuo.

“I don’t know.”

CHI: Aks im!

“Ask him!”

51) CHI: We i kii de?

“Where is the key?”

INV: Si yo av i.

“See, you have it.”

CHI: Oo.

“Oh.”

INV: Mmhmm.

“Mmhm.”

CHI: Fi yo kii de?

“(Where) is your key?”

INV: A fi mi kii dat.

“That key is mine.”

CHI: Ø Ø get i fram? (KEM 2;07)

Ø Ø get it from


181

“(Where did you) get it?”

INV: mm?

“mm?”

CHI: Get i fram? Mm?

“(Where did you) get it, mm?”

INV: We mi get i fram?

“Where did I get it?”

CHI: Yes.

“yes.”

INV: Mi get i fram out a di kyar.

“I got it from out the car.”

52) CHI: We yo a du an tamir fuon?

“What are you doing on Tamir’s phone?”

NIC: Aa mi no nuo.

“Aaa I don’t know.”

CHI: Aks tamir.

“Ask Tamir.”

NIC: Yuu aks tamir we mi a du pan i fuon.

“You ask Tamir what am I doing on the phone.”

CHI: Ø Ø aa du pan i fuon? (ALA 2;08)

Ø Ø PROG do on DET phone

“(What is she) doing on the phone?”

53) CHI: Ø Ø niem agen? (SHU 2,03)


182

Ø Ø name again?

“(What’s his) name again?”

INV: Mi no nuo.

“I don’t know.”

NIC: Aks yo mada we im niem kaa wi no nuo we im niem, gwaan go aks yo

mada, aks momi we im niem.

“Ask your mother what’s his name because we don’t know what his

name is, go and ask your mother, ask Mommy what’s his name.”

CHI: Momi Ø Ø niem? (SHU 2,03)

Mommy Ø Ø name

“Mommy, (what’s his) name?”

MOT: Ronel.

“Ronel.”

54) CHI: Grampaa kova op im mout.

“Grandpa covers up his mouth.”

INV: Grampaa kova op im mout agen.

“Grandpa covers up his mouth again.”

CHI: Ø Ø kova op im mout agen fa? (COL 2;08)

Ø Ø cover up 3SG mouth again for

“(Why did he) cover his mouth again?”

In addition to the context of utterance and the intonation, evidence for null wh utterances

comes from the stranding of the preposition fa. As seen in (54) above, in JC, ‘why’ may be

expressed as wa… fa “what… for”. Our data reveals that there are 32 of these “what… for”


183

questions for which the wh-element is omitted and leaving fa stranded, of which 15 also omits

the subject. In these examples, the stranding of the preposition is clear evidence for a null wh.

Also we note that there is the presence of a ‘question semantic unit (QSU)’27

in some of the

wh-constructions, for example mek in (55):

55) Ø mek it a work so? (ALA 2;08)

(Why) make 3SG PROG work so

“Why is it working like that?”

This QSU must occur with a wh-word. The lone occurrence of the QSU provides direct

evidence for a null wh-element.

Where wh-elements are dropped, declaratives can be distinguished from interrogatives, or wh-

interrogatives from yes/no’s, by means of context, intonation and the occurrence of the QSU.

Examples (56) – (58) demonstrate discourse contexts where the same phrase im gaan, by the

same informant, is interpreted as wh-question, declarative and yes-no question respectively.

56) CHI: Ø im gaan? (RJU 2;06)

Ø 3SG gone

“Where did he go?”

INV: mi no nuo we im gaan.

“I don’t know where he went.”

27 As proposed by Muysken and Smith (1990) a QSU gives additional information regarding what is questioned. It is a part of the adult grammar. No material can intervene between the wh-word and the QSU. Also, with the exception of mek ‘make’ in wa mek ‘why’ (literally ‘what make’) the QSU cannot occur on its own without the wh-word as a wh-question marker.


184

57) MOT: kaal joshwa, kaal joshwa no.

“Call Joshua, call Joshua (won’t you?)”

CHI: im gaan. (RJU

2;07)

3SG gone

“He is gone.”

58) MOT: si wan botaflai, sii im ova dier so, yo sii im? Sii im de.

“There’s a butterfly, see it’s over there, do you see it? There it is.”

CHI: mm, im gaan? (RJU 2;03)

COM28

3SG gone

‘Mm, is it gone?’

MOT: yes im gaan, an im priti.

“Yes it’s gone, and it was pretty.”

7.5.3 Null Subjects and Yes/No Questions in JC

We can now turn to yes/no questions and determine to what extent this construction is

consistent with null subjects in child JC. Of the 2501 interrogative utterance in our corpus

during the stage where null subjects are robustly attested, 1153 (46.1%) are yes/no questions,

as exemplified in (59) to (62):

59) Yuu bai i? (KEM 2;06)

2SG buy it

“Did you buy it?”

28 Communicator (COM) is used to refer to utterances that bear no syntactic content.


185

60) Yo mek im paas? (ALA 2;05)

2SG make 3SG pass

“Did you allow him to pass?”

61) Im a wiet pahn yo? (COL 2;08)

3SG PROG wait on 2SG

“Is he waiting on you?”

62) Yaa kil im? (RJU 2;03)

2SG~PROG kill 3SG

“Are you killing it?”

We observed that subjects can be freely dropped with yes/no questions as exemplified in (63)-

(66).

63) Ø Ø jrap? (TYA 2;01)

Ø Ø drop

“Am I going to fall?”

64) Ø Ø rait pan buk? (SHU 2;02)

Ø Ø write on book

“Am I to write on the book?”

65) Ø Ø bon yo? (KEM 2;02)

Ø Ø burn 2SG

“Will it burn you?”


186

66) Ø kyaahn kom aaf ? (RJU 2;03)

Ø can’t come off

“Can’t it come off?”

As displayed in Table 18, of the 1153 yes/no questions involving a verbal element, a total of

542 (47%) have omitted the subject. Thus, null subjects in yes/no questions are comparative

to that found in declarative utterances, which have 36.7% (4574 of 12463) null subjects.



1;8.0 0 0

0 0

1;8.5 0 1

0 1

1;9.0 1 1

1 1

1;9.5 0 1 0 5

0 6

1;10.0 0 0 0 0 0 0

0 0

1;10.5 1 0 1 15 0 0

2 15

1;11.0 0 0 0 11 0 0

0 11

1;11.5 0 1 0 6 1 1 0 0

1 8

2;0.0 3 1 0 11 0 0 0 2

3 14

2;0.5 3 2 0 2 0 1 0 1 0 2

3 8

2;1.0 1 4 0 2 0 5 0 0 0 5

1 16

2;1.5 3 0 1 7 3 3 0 1 1 2 2 1 10 14

2;2.0 1 2 4 30 3 8 0 0 0 3 6 2 14 45

2;2.5 0 1 8 26 1 2 0 0 0 14 3 1 12 44

2;3.0 0 3 3 4 3 2 0 0 0 13 3 7 9 29

2;3.5 1 2 5 22 2 4 0 0 0 6 2 6 10 40

2;4.0 2 3 5 11 7 7 0 0 5 7 2 0 21 28

2;4.5 0 0 7 7 1 1 0 0 2 23 3 0 13 31

2;5.0 2 0 8 9 2 3 0 0 1 9 5 1 18 22

2;5.5 5 3 15 5 13 7 0 0 2 11 14 4 49 30

2;6.0 0 0 1 1 5 5 0 0 4 4 11 3 21 13

2;6.5 3 2 8 0 15 5 1 1 2 8 10 3 39 19

2;7.0 6 4 17 2 13 3 0 0 4 4 6 11 46 24

2;7.5 3 1 8 1 13 3 0 0 1 10 8 13 33 28

2;8.0 5 3 28 8 9 3 1 2 4 5 5 3 52 24

2;8.5 14 3 25 4 9 11 1 0 2 2 4 0 55 20

2;9.0 2 0 11 1 3 6 0 0 8 4 6 1 30 12

2;9.5 3 1 19 2 6 4 0 0 4 0 6 4 38 11

2;10.0 7 3 21 1 5 1 0 0 1 1 2 2 36 8

2;10.5 13 5 16 2 0 0 0 0 1 0 0 0 30 7

2;11.0 8 1 7 3 3 1 0 0 14 2 7 0 39 7

2;11.5 10 3 2 2 3 0 6 0 4 1 25 6

TOTAL 87 48 228 201 119 88 6 7 62 135 109 63 611 542

Table 18: Actual Production of Null subjects with Yes/No Questions


187

Additionally, like in the production of null subjects with declaratives, we observe a gradual

decline in the production of null subjects with yes/no questions, as shown in Figure 9.

Figure 9: Percentage production of Null Subjects in Yes/No Questions

This strong parallel between declaratives and yes/no questions should be accounted for,

likewise the robust attestation of null subjects in phrases having an unpronounced +Q

element. We therefore, in the next section, sketched out a revised approach to the Privilege of

the Root to account for these findings.

7.6. REVISING THE PRIVILEGE OF THE ROOT MECHANISM.

The pattern that emerges from the previous section is therefore the following:

67) a. Overt wh overt subject: ok

b. Null wh overt subject: ok

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

AGE (mths)

% PRODUCTION


188

c. Yes/No overt subject: ok

d. Overt wh null subject: *

e. Null wh null subject: ok

f. Yes/No null subject: ok

While cases (67)a-b-c-d are predicted, cases (67)e-f are unexpected under the traditional view

of Truncation. Questions with null wh-elements such as (45)-(48) plausibly involve

movement of the wh-element to a designated landing site in the left periphery, much as

ordinary constituent questions, except that the wh-element is not pronounced. Straightforward

evidence for a movement analysis is offered by the V1 shape of such questions in V2

languages (with the null wh-element occupying the initial position and satisfying the V2

constraint), and by the absence of the phenomenon in in-situ configurations (Yamakoshi

2002). In addition, direct evidence for a movement analysis in JC comes from the observation

of a change in prepositional form (fi/fa) in wh phrases, depending on whether the preposition

is followed by a wh trace or an overt object (Durrleman-Tame 2008), or some other overt

element. As shown in examples (68) and (69) below, fa is used in utterances that are followed

by a trace indicating movement of the wh-element (68), while fi is used in utterances where

the wh-element remains in situ for example in echo questions (69).

68) Wa yu put i aan fa/*fi?

“Why do you put it on?”

69) Yo put i aan fi/*fa wa?

“You put it on for what?”

As mentioned earlier the stranding of fa in null constituent questions, as in (70) below,

indicates that there is a wh trace, which provides evidence for movement of the wh-element.


189

Note that if the null wh-element were to remain in situ, we would have expected fi as in (69),

none of which are in our corpus.

70) Ø Ø put i aan fa? (ALA, 2;03)

Ø Ø put it on for

“(Why do you) put it on?”

We therefore assume a movement analysis for the JC cases, which straightforwardly accounts

for the formal and interpretive properties of these structures as constituent questions, and for

the option of a null wh-element as a particular case of the Privilege of the Root. But then, if

(71) has a representation like (72)

71) Ø Ø get i fram? (KEM, 2;07)

Ø Ø get it from

72) wherenull C younull get it from __

clearly the CP structure is needed to permit movement of the null where, hence the structure

cannot be truncated at the IP level, and the subject cannot be the Spec of the root. Still a null

subject is possible in this environment. Likewise for Yes/No questions, such questions do not

correspond to overt morphemes in the target language, however they are complete Force

phrases, with the interrogative force explicitly expressed in a dedicated position in the left

periphery, requiring projection of the CP layer. But again, a null subject is permitted in this

environment. The approach to the Privilege of the Root based on Truncation qua radical

absence of structure must therefore be revised to accommodate these cases.


190

In order to sketch out such a revision, we would like to build on the idea that the Privilege of

the Root can be naturally expressed through the spell-out mechanism associated to Phase

Theory (Chomsky 2001, Nissenbaum 2000), as proposed in Rizzi (2005, 2006). In order to

make successive-cyclic movement out of a phase possible, Phase Theory must assume that

when a head defining a phase is reached by the computation, only the complement of the

phase head is sent to spell-out, not the whole phase. This allows the material in the phase edge

to remain available for further computation, e.g., further movement. The approach raises the

question of how the edge of the root clause (e.g., a wh-element or a topic in a main clause) is

ever sent to spell-out. Rizzi’s proposal (building on Chomsky’s and Nissenbaum’s ideas)

capitalizes on the cartography of the higher part of the clausal structure. Assume that the

structural map is roughly like the following sequence (Rizzi 1997, Rizzi and Shlonsky 2007

and much subsequent work):

73) Force Top Foc Fin Subj …..T …

If Force is the phase defining head, a wh-element (moved to Spec Foc), a topic, etc will

normally be spelled out, as they are part of the complement of the Force head. What about the

Privilege of the Root? The proposal in Rizzi (2005, 2006) is that the choice of the category

counting as the root phase may be parametrized in part: a Topic Drop language (such as

colloquial German) would involve the selection of TopP as a possible root phase, so that a

topic in the TopP head could be left not spelled out. A wh-drop language would involve the

selection of FocP (if indeed FocP is the normal landing site of wh-movement) as a possible

root phase, so that a wh-element in its edge could remain unpronounced; a root subject drop

language (such as certain registers of English, Haegeman 2000, etc.) would involve the choice

of SubjP as a possible root phase, so that a root null subject would be possible, etc. Child

grammars are assumed, in this system, to systematically recruit such UG-consistent

parametric options, which permit a maximization of clause-initial ellipsis; such options are


191

later abandoned, around the end of the third year of life, if they are not target-consistent,

hence if they are not supported by experience.

This approach, much as earlier versions of the Truncation idea, does not capture case (67)e-f,

though: if the root phase head here is SubjP, there would be no room for further wh-

movement. A similar problem arises for the variant presented in Wexler (2014), in which

Truncation at the IP level is made contingent on certain scope-discourse properties of the

structure. Clearly, in order to capture (67)e-f, the computation must be allowed to continue, at

least a little bit, on layers higher than the spelled-out material.

The empirical problem raised by the simultaneous occurrence of null +Q elements and null

subjects suggests a revision of the characterization of the Privilege of the Root provided by

Truncation in the traditional sense: perhaps, the special property of the root should not be

expressed in the definition of phase nodes, but in the functioning of the spell-out mechanism.

What is special about the root environment may not be the fact that different categories could

count as root phases, but rather the fact that the spell-out mechanism may have more options

than in non-root environments, so that it may be possible to “pronounce less” in root

environments, in otherwise uniform structural representations. In the revised system, the

clausal phase node would uniformly be the Force head, while the spell-out mechanism could

be based (for a first approximation) on the following scheme:

74) Spell out α, where α is c-commanded by the root Force

α can always be the complement of Force, as in run-of-the-mill main clauses and in all

embedded environments; but in root environments α could parametrically assume different

values: topic drop languages could have α = the node immediately under TopP, root subject

drop languages could have α = the node immediately under SubjP, etc. The structure higher


192

than the spell-out point α could involve further computation (external and internal merge), but

it would be left unpronounced. Child systems would recruit all such parametric options

making possible a maximization of ellipsis at the root of the clause.

The revised conception of the Privilege of the Root in terms of selective spellout captures all

the cases of (67). In (67)a, α can be the complement of Force, an option which is always

available in child and adult grammars. In (67)b-c, α may be SubjP: the whole content of

SubjP, including the subject pronoun, is spelled out here, while the whole CP zone, including

the +Q element, is null; this captures the null wh phenomenon and Yes/No questions. In

(67)e-f, α is a node under SubjP, say TP in (73): both the +Q element and the subject are

higher than the spell-out point, so that they are both null; but the computation has continued in

the higher layers, involving, among other things, wh-movement. This may account for the

strong parallelism of null elements in Yes/No questions and declaratives: As in Yes/No

questions, α may also be lower than SubjP in declarative clauses, hence permitting null

subjects. As there are no overt morpheme corresponding to declarative force in the left

periphery, the computation is allowed to continue pass the spell-out point to integrate SubjP

and the whole left periphery, including the appropriate force marker of declarative phrases.

And (67)d continues to be excluded, as in the structure-building approach to Truncation: the

fact that the wh-element in the C-system is spelled out indicates that the spell-out point must

be higher in this structure, at least as high as the FocP; but then the SubjP layer is lower than

the spell-out point (73), and therefore the subject must be pronounced. The whole paradigm

(67) is thus captured.


193

7.7. SUMMARY AND CONCLUSION

The following table summarizes our findings on the distribution of null and overt subjects in

four crucial environments in child JC: declarative clauses, constituent questions with an overt

wh-element, yes-no questions, and constituent questions with a null wh-element.

Figure 10: Distribution of null subject

What immediately sticks out from this figure is the virtual absence of null subjects in the

environment following an overt wh-element. This fact, and the sharp contrast with the robust

attestation of null subjects in declaratives, gives strong support to the view that early null

subjects are cases of root subject drop, a manifestation of the more general phenomenon

63.3%

36.7%

53.0%

47.0%

98.9%

1.1%

77.4%

22.6%

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%


194

dubbed “the Privilege of the Root”, permitting an exceptional freedom in not pronouncing the

initial edge of the clause, with distinct manifestations in adult and child systems (Rizzi 1992,

2006). This conclusion is also supported by the contrast between a high rate of (initial) subject

drop and a low rate of obligatory object drop (section 7.2). In addition to constituent questions

with an overt wh-element, child JC, much as other child languages, also manifests sizable

numbers of constituent questions which appear to involve a null wh-element, plausibly

another manifestation of the Privilege of the Root. If the virtual absence of subject omission

in overt post-wh environments is immediately predicted by the traditional Truncation

approach, expressed in terms of radical absence of the higher layers of the clausal structure,

the significant attestation of null subjects after questions plausibly involving a null +Q

element (4th and 8th bars in figure 10) is not: in such cases, the syntactic computation

(movement of the null wh-element in the cases of null constituent questions) must be able to

target a position higher than the position of the null subject. The generalization remains

correct, though, that early null subjects cannot be preceded by an overt element. This

observation led us to explore a revision of the Truncation approach in terms of the spell-out

mechanism, rather than of the structure-building operations. In terms of phase theory, the

head of the root phase triggers spell-out more liberally than embedded phase heads, allowing

the initial chunk of the clause to remain unpronounced: this is consistent with more

unpronounced structural layers in the initial part of the structure, as in the case of null wh-

elements followed by null subjects. The spell-out approach to Truncation and the Privilege of

the Root are also consistent with the high proportion of null subjects found in declaratives in

child JC.

Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization

195

CHAPTER 8

THE ACQUISITION OF WH-INTERROGATION, FOCALIZATION AND

TOPICALIZATION

8.1 INTRODUCTION

In this chapter, I explore the development of wh-questions, focus and topics. The acquisition

of these three phenomena is usually viewed as involving quite complex movement operations.

By looking at these constructions, we will investigate the emergence of left-peripheral

movement in children’s grammar. We will see that from their early productions children

acquiring JC are knowledgeable of the universal constraints governing syntactic movement

and ordering.

Cross-linguistically, there are varying strategies in the formation of wh-interrogative

utterances, focus and topics. For example, some languages employ options where the wh-

element remains in situ while in others it must undergo movement to the left periphery of the

clausal structure. Likewise for focus and topics in some languages they are signaled by left-

peripheral syntactic movements while in others the distinctions are associated to prosodic

prominences. Additionally, some languages require the presence of a marker in instances of

interrogation, focalization and topicalization. This cross-linguistic variation in the formation

of wh-interrogative, focalized and topicalized utterances may contribute to complicating the

acquisition of these structures (in line with Costa & Szendroi (2006) for focus marking) since

children will have to find out which strategy is used in their language. It is therefore expected

that they will make mistakes in the course of their development.


196

The chapter is organized as follows: Section 8.2 studies the details of wh-question formation

in the corpus. In section 8.3 we examine constituents that are overtly marked for focalization,

concentrating primarily on those involving movement to the left periphery of the clause. In

addition we examine utterances which are focalized, but for which the overt focus marker is

omitted. Section 8.4 discusses topicalization zooming in on predicative utterances involving

leftward movement and shows where children produce utterances resembling left dislocated

topics. Section 8.5 provides evidence for cartographic ordering of the left periphery in JC. The

chapter ends with a short discussion of the main conclusions.

8.2 ACQUISITION OF INTERROGATION

The general cross-linguistic consensus in the literature is that there is a relatively robust

sequence of acquisition of wh-questions, in which questions that encode syntactically simple

relations, for example what, where and who are acquired before wh-questions that refer to

more complex concepts, such as why, how and when (for English: Bloom, Merkin & Wootten

1982; for Serbo-Croatian: Savic 1975; for Korean: Clancy 1989; for German: Wode 1975,

etc). This section aims to establish the extent to which the predicted order of acquisition of

wh-questions applies to children acquiring JC, and to propose a syntactic analysis for the

attested developmental pattern.

8.2.1 Acquisition of Constituent Questions in JC

The data, as presented in Table 1, reveal that the informants produced a total of 3196 wh-

questions during the period starting from 20 months to 40 months of age. They primarily

produced ‘where’, ‘what’ and ‘who’-type questions. Together, these three question types

constitute 85.3% of the wh-questions in the entire corpus.


197

AGE Wh-questions TOTAL

(MTHS) what where who why how when which

1;8.0 0 0 0 0 0 0 0 0

1;8.5 0 0 0 1 0 0 0 1

1;9.0 0 1 1 0 0 0 0 2

1;9.5 1 1 4 0 0 0 0 6

1;10.0 2 1 0 0 0 0 0 3

1;10.5 10 2 2 0 0 0 0 14

1;11.0 2 2 0 0 1 0 0 5

1;11.5 1 11 0 0 0 0 0 12

2;0.0 4 16 8 0 0 0 0 28

2;0.5 3 6 9 0 1 0 0 19

2;1.0 15 8 5 0 0 0 0 28

2;1.5 14 10 11 0 0 0 0 35

2;2.0 30 17 35 0 0 0 0 83

2;2.5 26 14 40 0 1 0 0 81

2;3.0 7 14 28 1 1 0 0 51

2;3.5 20 28 17 1 1 0 0 67

2;4.0 13 14 7 13 0 0 0 47

2;4.5 21 16 10 44 1 0 0 92

2;5.0 14 11 11 4 0 0 0 40

2;5.5 72 39 24 21 1 0 0 157

2;6.0 8 39 12 6 0 0 1 66

2;6.5 56 45 11 7 0 0 0 119

2;7.0 56 80 33 10 0 0 0 179

2;7.5 57 62 22 22 1 0 0 164

2;8.0 76 44 31 8 4 0 1 164

2;8.5 34 33 24 5 3 0 0 99

2;9.0 48 63 32 26 4 0 0 173

2;9.5 48 46 41 37 2 0 0 174

2;10.0 98 37 23 33 2 0 1 194

2;10.5 27 27 17 7 1 0 0 79

2;11.0 24 65 14 15 6 2 0 126

2;11.5 17 25 14 4 2 0 0 64

3;0.0 54 34 21 14 3 0 3 129

3;0.5 36 48 20 17 5 0 0 126

3;1.0 40 30 16 8 6 0 1 101

3;1.5 18 14 20 14 1 0 1 68

3;2.0 23 38 20 33 10 0 1 125

3;2.5 22 7 6 2 0 0 1 38

3;3.0 19 36 7 4 7 0 0 73

3;3.5 11 2 4 1 1 0 0 19

3;4.0 13 14 10 6 8 0 0 51

3;4.5 33 29 15 4 16 1 0 98

TOTAL 1072 1029 625 368 89 3 10 3196

Table 1: Production of Wh-questions with age


198

The order of first attestation of the questions showed that a why-question was the first wh-

question to be produced at 20 months of age. This is not in keeping with expectations as

presented in the literature. Closer analysis of this utterance revealed that it was the single

sporadic utterance of a why-question in COL’s production for the entire period under

examination. Its use is therefore non-productive and can be disregarded for the present

analysis. We note that who, where and what questions were all produced during the 21st

month of life. How and why questions were used productively much later, and when and

which questions occurred rarely, even at 40 months when the study ended. Table 2 reveals

that what, where and who questions have been produced during the single word stage, why

and how during the initial combinations, and when and which during the later multiword

combination stage.

MLU what where who why how when which

< 1.49 3 5 5 0 0 0 0

1.5 - 2.49 58 82 25 1 4 0 0

2.5 - 3.49 329 247 235 69 3 0 1

3.5 - 4.49 391 331 192 129 11 2 1

4.5 > 288 361 164 168 71 1 8

Table 2: Production of Wh-questions with MLU

With regards to what and where questions, there is a strong favor for object questions, and

subject questions with who questions as exemplified in (1) – (3) respectively.

1) We kyat a liedong pan? (TYA 2;09)

What cat PROG lie on

“What is the cat lying on?”

2) We yu fren de? (SHU 2;05)


199

Where 2SG friend LOC

“Where is your friend?”

3) Uu ina i kyaar? (RJU 2;07)

Who into DET car

“Who is in the car?”

According to Van Valin (1998), for English, subject questions are expected to be produced

first as they do not involve subject auxiliary inversions and are as such less complex. This was

not borne out by his data however. Interestingly also, we would not expect this to hold in JC,

as, consistent with the target system, there are no inversions or insertions of auxiliaries or

‘dummy do’s’ throughout the data, for subject nor object questions. Both question types have

been produced at the same time.

The data provide additional evidence for the sequence in which early wh-forms are produced,

and arguably reflect the relative syntactic complexity among the different wh-forms. The

pronominal forms what, where and who were learnt before the sentential forms why, how,

when and which, as reported in previous studies of first language acquisition.

8.2.2 In-situ Wh-Phrases

The data reveal 27 in-situ wh-questions, 11 of which were echo questions as in (4):

4) INV: Yo afi go aks granmaa fi mek som fi yo leeta.

“You have to ask grandma to make some for you later.”


200

CHI: Aks uu? (ALA 2;04)

Ask who

“Ask who?”

The other 16 in-situ utterances appear to be target-consistent questions where the in-situ-wh is

permitted in ‘discourse-bound’ restricted contexts, as exemplified in (5) and (6).

5) CHI: We i bi?

What 3SG be

“What is it?”

INV: Rekaada.

“Recorder”

CHI: Rekaada?

recorder

“Recorder?”

INV: Ye.

“Yes”

CHI: Fi rekaad uu? (SHU 2;05)

INF record who

“To record who?”

6) CHI: Aa uol i fi yo.

1SG-PROG hold 3SG for 2SG

“I am holding it for you”

CUZ: Im se im gaa uol i fi yo.

“He said he is going to hold it for you”


201

INV: Ye a fi yo uon, mi a gi yo bot mi did a_go elp yo put aan i, yo no

wahn no elp? mm?

“Yes it is yours, I am giving it to you but I was going to help

you to put it on, don’t you want any help? Mm?

CHI: Do wat? (RJU 2;01)

Do what

“To do what?”

The data reveal an interesting pattern in development of wh-questions: the first questions

contained only fronted wh-elements. In-situ utterances were produced later, the first of which

occurred during the period when MLU is at 2.5 – 3.49. This pattern is not universal as other

languages, such as French (Hamann, 2006), show that wh-in-situ is the first wh-question to

emerge. It is suggested that in-situ wh-questions are mastered first by children as they are

more economical and do not involve overt movement of the wh-word (see Kampen 1997,

Jakubowicz and Strik 2008, Hamann 2006, Rizzi 2000). We therefore need to account for its

late development in JC. I tentatively propose that children acquiring JC rarely explore the in-

situ option in wh-question formation as it rarely exists in the target language. Whenever it is

employed in the adult language, it is most predominantly as an echo-question, as in (4) above.

Given that it is not frequent in the input could possibly account for its late development in

child’s speech.

8.2.3 Movement in Early Wh-Questions in JC

Wh-phrases in JC are positioned in the specifier of CP. They must be moved from their base

position in as in (7a and b), and undergo certain transformations in order to yield the fronted


202

interrogative in (7c). These transformations may be motivated on the grounds of the well-

formedness condition on question-formation referred to as the Wh-Criterion (Rizzi 1996).

This criterion stipulates that a wh-operator and the head carrying the wh-feature must be in

specifier-head relationship.

7) a. Dis luk laik wahn apl.

DEM look like DET apple

“This looks like an apple”

b. Dis luk laik wa?

DEM look like what

“This looks like what”

c. Wai dis luk laik ti? (ALA 2;07)

what DEM look like

“What does this look like?”

The following examples demonstrate overt fronting of the wh-operator in the corpus along the

lines presented in (7) above:

8) Wai im niem ti? (COL 2;06)

what 3SG name

“What is his name?”

9) Wei i baal de ti? (KEM 2:04)

where DET ball LOC


203

“Where is the ball?”

10) Wei yaa go ti? (TYA 2;06)

where 2SG~PROG go


11) Wei shi du dat fa ti? (RJU 2;04)

what 3SG do that for

“Why did she do that?”

12) Oui yu lak i ti? (SHU 2;09)

how 2SG lock 3SG

“How did you lock it?”

If the wh-criterion were inoperative in early JC, we would expect children to produce wh-

questions in which the wh-element had not moved to clause initial position. As detailed in the

previous chapter, direct evidence for a movement analysis in JC also comes from the

observation of a change in prepositional form (fi/fa) in wh phrases, depending on if the

preposition is followed by a wh trace or an overt object (Durrleman-Tame 2008), or some

other overt element. As shown in examples (13) - (16) below, fa is used in utterances that are

followed by a trace indicating movement of the wh-element, while fi is used in utterances

where the wh-element remains in situ for example in echo questions as seen in (17).

13) Wei yu put aan i fa ti? (ALA 2;04)

What 2SG put on 3SG for


204

“Why do you put it on?”

14) Wei yaa jraiv mi kyaar fa ti? (KEM 2;11)

what 2SG~PROG drive my car for

“Why are you driving my car?

15) Wei yo opin i fa ti? (RJU 2;11)

what 2SG open 3SG for

“Why did you open it?”

16) Wei yo tikl mi fa ti? (SHU 2;05)

what 2SG tickle 1SG for

“Why did you tickle me?”

17) Ø ina i ous fi wa? (KEM 2;09)

Ø into DET house for what?

“Why is he in the house?”

The distinctions in the use of fi/fa reveal that children are aware of the movement rules with

respect to question formation from an early age. We can therefore conclude that children

acquiring JC are knowledgeable of the syntactic constraints governing question formation, as

required by the target system. This conclusion is upheld by analyses of questions in early

Italian (Guasti 2000), Swedish (Santelmann 1997), Dutch (Haegeman 1995a), German

(Clahsen, Kursawe and Penke 1995), etc.


205

8.2.4 Errors in Production

The literature reports that children, acquiring English, make errors in producing wh-questions,

the most common of which are failure to produce an auxiliary and the incorrect placement of

an auxiliary after, rather than before the subject (Guasti 2000; Brown 1973; Bellugi 1971;

Guasti & Rizzi 1996; etc.). As adult JC does not employ insertions or inversions of an

auxiliary, children acquiring JC made no errors in this regard and were therefore target-

consistent from the start.

We will now examine the acquisition of focus in JC.

8.3 ACQUISITION OF FOCUS IN JC

According to Durrleman-Tame (2008), adult JC sentences involving focalization generally

place the focalized constituent at the front of the sentence. It is necessary for this constituent

to be immediately preceded by an overt focus particle a which marks focalization as in (18):

18) A domplin mi a kuk. (RJU 2;11)

FOC dumplings 1SG PROG cook

“It is dumplings that I am cooking.”

Focalization in this language is non-recursive, and importantly, cannot involve a resumptive

pronoun. Durrleman-Tame (2008) shows where focus preposing in JC can involve a variety of

categories, such as determiner phrases, adjectival phrases and verbal phrases. Focalized

adjectival and verbal elements, while in the fronted position, seem to bear what can be


206

analyzed as nominal traits, suggesting that this focus position in JC is a nominal position (see

Durrleman-Tame, 2008 for discussion).

The data reveal that focalization, a complex phenomenon involving movement to the left

periphery (Durrleman and Shlonsky, 2015) is evident from very early in the production of the

children. The first use of the focus marker ‘a’ was seen at 21 months of age:

19) A biebi. (COL 1;09)

FOC baby

“It is the BABY’s.”

20) A uu? (ALA 1;09)

FOC who

“Who is it?”

The early use of the focus particle a is not totally unexpected as similar reports of early use of

focus has been documented for other languages. For example, as revealed from longitudinal

spontaneous speech data, the production of the overt focus particle ‘auch’ is among the first

lexical items acquired by German learning children (e.g. Nederstigt 2001; Penner, Tracy &

Weissenborn 2000). It is used basically adultlike, prosodically, syntactically, and

semantically. More specifically, Nederstigt (2001) found that children as young as 1 year and

6 months are sensitive to the relation between the particle ‘auch’ and the focus of the

sentence, as marked by intonation in their productions of focus constructions.


207

Focalization was noted to appear early in both declarative and interrogative phrases, both in

subject position (exemplified in (21) – (26)) and in predicate positions.

21) A Maruun dwiit. (KEM 2;07)

FOC Maroon do~it

“It is Maroon who did it.”

22) A yuu bad. (SHU 2;04)

FOC 2SG bad

“It is you who is bad”

23) A mi piipi op in de. (TYA 2;08)

FOC 1SG urinate up in LOC

“It is me who urinated in there.”

24) A uu kaal yo? (ALA 2;08)

FOC who call 2SG

“Who is it that called you?”

25) A uu bai dis? (COL 2;08)

FOC who buy DEM

“Who is it that bought this?”

26) A uu mash op i? (RJU 2;08)

FOC who mash up it


208

“Who is it that destroyed it?”

For the present study we will be concentrating on the production of predicative focus in which

the focalized constituent is fronted. According to Durrleman-Tame (2008) these syntactically

focalized elements obligatorily leave a gap in its base position in JC, as evidenced by the lack

of a resumptive clitic. This will be demonstrated in the examples provided throughout the

remainder of the discussion on focalization. Table 3 details the distribution of the production

of the overt focus marker with fronted predicative constituents in the corpus:


209

AGE Total -subj WH OBJ VERB PP ADV

1;10.5 4 4 0 0 0 0

1;11.0 1 1 0 0 0 0

1;11.5 0 0 0 0 0 0

2;0.0 4 3 0 0 1 0

2;0.5 1 1 0 0 0 0

2;1.0 1 1 0 0 0 0

2;1.5 4 4 0 0 0 0

2;2.0 13 13 0 0 0 0

2;2.5 21 20 1 0 0 0

2;3.0 1 1 0 0 0 0

2;3.5 19 14 3 1 1 0

2;4.0 4 3 1 0 0 0

2;4.5 5 4 0 0 1 0

2;5.0 1 0 0 0 1 0

2;5.5 4 3 1 0 0 0

2;6.0 4 0 4 0 0 0

2;6.5 7 4 3 0 0 0

2;7.0 11 8 2 0 0 1

2;7.5 6 3 3 0 0 0

2;8.0 24 17 3 3 0 1

2;8.5 7 3 3 0 0 1

2;9.0 17 11 3 0 0 3

2;9.5 22 14 6 0 1 1

2;10.0 17 12 4 0 1 0

2;10.5 14 4 8 0 1 1

2;11.0 9 3 3 1 0 2

2;11.5 21 8 11 1 1 0

3;0.0 26 19 4 0 0 3

3;0.5 18 13 5 0 0 0

3;1.0 17 11 4 0 0 2

3;1.5 12 8 3 0 1 0

3;2.0 26 18 3 0 0 5

3;2.5 23 8 5 0 0 10

3;3.0 16 7 6 0 0 3

3;3.5 12 5 7 0 0 0

3;4.0 14 11 0 0 1 2

3;4.5 32 29 2 0 0 1

Total 438 288 98 6 10 36

Table 3: Production of Overt Predicative Focalized Utterances


210

During the period under examination, we saw 43829

fronted focalized predicative utterances

overtly marked with the focus particle a. This represents 13% of the total focalized utterances

in the corpus (3341 utterances). Children’s production appears to be mostly adult-like, thereby

in keeping with the target language. Children always produce the morpheme a sentence

initially, indicating that its argument to the right is focalized. There is general consensus that

the type of movement involved in fronting the focalized argument (or predicate) in Atlantic

Creoles is similar to the wh-type movement found in wh-questions and relative clauses

(Winford 2008). The same restrictions governing the change in prepositional form fi/fa which

is evident in cases of wh-movement as discussed in section 8.2.3 above is found in focalized

constructions (Durrleman-Tame 2008). Note that fi is not allowed in utterances which

undergoes left-peripheral focus movement:

27) A Kyahri mi a luk fa. (RJU 2;11)

FOC Kyahri 1SG PROG look for

“It is Kyahri that I am looking for.”

28) Mi a_go luk fi Kyahri. (KEM 2;09)

1SG PROS look for Kyahri

“I am going to look for Kyahri.”

29

Note that utterances involving predicative focalization but for which the subjects are expletives (and need not be overtly pronounced in JC) were not included in this analysis. For example: i) INV: A we dadi kyari? “What is it that daddy carried?” CHI: A skeliyan. (COL 2;03) FOC escallion “It is escallion.” Only utterances requiring the overt phonological realization of the subject was included in this analysis.


211

Despite the complexities associated with movements, the data reveals that a is attested to the

left of non-subject wh-pronouns in constituent questions from an early age of 22 months. A

total of 288 non-subject focalized wh- interrogatives were produced as exemplified in (29) –

(34). This represents 69% of the focalized wh-questions in the corpus.

29) A we Kyahri de? (KEM 2;09)

FOC where Kyahri LOC

“Where is it that Kyahri is?”

30) A ou yu tek i aaf? (SHU 3;03)

FOC how 2SG take it off

“How is it that you removed it?”

31) A we kola mi slipaz bi? (TYA 2;11)

FOC what colour 1SG slippers COP

“What colour is it that my slippers are?”

32) A we yo put i in de fa? (ALA 2;09)

FOC what 2SG put 3SG in LOC for

“Why is it that you put it in there?”

33) A wa Niki av? (COL 2;03)

FOC what Nicki have

“What is it that Nicki has?”


212

34) A we i baal? (RJU 2;08)

FOC where DET ball

“Where is it that the ball is?”

Also evident in the dataset was object focalization. Note that a cannot mark objects as

focalized unless they are moved from their base position to the left periphery of the clause.

Also note that there are no resumptive pronouns employed in these utterances. Of the 438

utterances under examination, fronted object focalization was demonstrated by 98 utterances,

as in examples (35) – (40) below.

35) A big manggo a av. (COL 2;07)

FOC big mango 1SG have

“It is a big mango that I have.”

36) A dengge dadi av. (ALA 2;08)

FOC dengue daddy have

“It is dengue that daddy has.”

37) A dadi mi se. (KEM 2;10)

FOC daddy 1SG say

“It is daddy that I said.”

38) A domplin mi a kuk. (RJU 2;11)

FOC dumplin 1SG PROG cook

“It is dumpling that I am cooking.”


213

39) A dat mi slippaz bi. (TYA 2;11)

FOC DEM 1SG slippers be

“That is what my slippers are.”

40) A nat mi biebi shi bi. (SHU 2;11)

FOC NEG 1SG baby 3SG be

“It is not my baby that she is.”

Interestingly we have also seen the realization of a copula-like element bi ‘be’ attested in

sentence final position with nominal predicates as in examples (39) and (40). These types of

utterances accounts for 34% of the fronted objects. The status of bi is discussed in the next

chapter, Section 9.6.

Overt focalization is also attested, but to a lesser extent, in other constructions. We have seen

6 cases where the verb was focalized, all of which are target-consistent. When verb predicates

are focalized the focus element also appears in situ as detailed in examples (41) – (46):

41) A toch a toch im. (COL 2;08)

FOC touch 1SG touch 3SG

“It is touch that I touched him”

42) A jag im mi a jag im. (KEM 2;11)

FOC drag 3SG 1SG PROG drag 3SG

“Pulling him is what I am doing”


214

43) A jrap yo jrap? (RJU 2;08)

FOC drop 2SG drop

“Is it that you fell?”

44) A nat riid mi a riid. (RJU 2;08)

FOC NEG read 1SG PROG read

“It is not reading that I am doing.”

45) A no kraab mi a kraab yo. (RJU 2;11)

FOC NEG scrape 1SG PROG scrape 2SG

“Scraping you is not what I am doing.”

46) A krach di graas a krach mi. (SHU 2;03)

FOC scratch DEF grass PROG scratch 1SG

“Itching me is what the grass is doing.”

These constructions appear to be V movement30

as opposed to VP movement. According to

Durrleman-Tame (2008), in X movement, the X reduplicates in the base position, while in

cases of XP movement, reduplication of the XP is banned. Aboh (2004) offers an alternative

perspective, claiming that the focalized V moves to a head focus position, as opposed to the

Spec FocP position that focalized XP occupies. The IP internal copy of the verb functions as a

resumptive verb or a copy in terms of Chomsky (1995). As demonstrated in examples (35) to

30

Except for example (42) where the focussed verb is shown to occur with its subject. The grammaticality of these types of utterances is questioned in Durrleman-Tame (2008).


215

(40), children do not reduplicate the fronted DP; thereby providing evidence of early

knowledge of the syntactic rules distinguishing between XP and X focusing31

.

Additionally, 10 prepositional phrases were focalized, none of which were target inconsistent,

as exemplified in (47) – (49):

47) A op de i jraiv. (COL 2;00)

FOC up LOC 3SG drive

“It is up there that it drove.”

48) A out a kuokanat Felisha out a kuokanat. (RJU 2;04)

FOC out PREP coconut Felisha out PREP coconut

“It is by the coconut (tree) that Felisha is.”

49) A op mi a go. (SHU 3;01)

FOC up 1SG PROG go

“It is up that I am going.”

Durrleman-Tame (2008) points out that only certain PPs can undergo focalization: those that

can occupy subject position and can be referred to by means of a pronoun, thus appearing to

have nominal status. As shown above, all the focalized PPs can be replaced by de-so ‘there’,

once again showing that children are aware of the syntactic rules.

31

Durrleman-Tame (2008) further points out that the focus position is a nominal position and as such focalized verbs are not really focalized V but rather a matrix of features corresponding to the root of the predicate (which assumes certain nominal traits).


216

With regard to the focalized adverbials, 36 utterances were detected in the corpus. 66% of

these utterances employed the adverbs so ‘so’:

50) A so yo jraiv it. (COL 2;07)

FOC so 2SG drive 3SG

“It is like this you drive it.”

51) A so yo jraa gorl? (TYA 3;02)

FOC so 2SG draw girl

“Is it like this you draw a girl? “

52) A so yo fool it. (RJU 2;08)

FOC so 2SG fold 3SG

“It is like this you fold it.”

Other types of adverbs are however focalized as presented in examples (53) – (55):

53) A onli wan buk yu kyari. (ALA 2;11)

FOC only one book 2SG carry

“It is only one book you carried.”

54) A im a stil im dat. (KEM 2;11)

FOC 3SG FOC still 3SG DEM

“It is him, that is still him.”


217

55) A yaso yu uopm it? (SHU 3;01)

FOC LOC 2SG open 3SG

“Is it here you opened it?”

A more detailed analysis of the data with respect to MLU as demonstrated in Table 4 reveals

that children have been productively producing overt focalized utterances from as early as

they start combining two words. This is however restricted to non-subject wh- constituent

questions, but nonetheless provides evidence of early non-subject focalization in JC. We see

that as children’s MLUs increase, the contexts in which focalized predicates occur widens.

Object DP focalization becomes productive during the stage where the first multiword

utterances are attested, and verbal, adjectival and prepositional focalization is evident.

MLU Total -Sub Wh OBJ VERB PP ADJ

< 1.59 1 1 0 0 0 0

1.5 – 2.49 19 18 0 0 1 0

2.5 - 3.49 81 63 13 1 3 1

3.5 - 4.49 92 66 20 4 0 2

4.5 > 245 140 65 1 6 33

Table 4: Production of overt focalized utterances with MLU

8.3.1 Overt Focalized Fronted Constituents and Null Subjects

Recall that in the previous chapter, null subjects are robustly attested in the utterances of

children learning JC, however they are restricted to clause initial positions and are virtually

absent from instances of wh-preposing. Based on the need of the overt CP to hold the wh-

element, Truncation is not possible. In these constructions, the subject is no longer the Spec of

the root and cannot be dropped. Likewise, in utterances with left-peripherical movement of a


218

focalized constituent, CP is required to host the focalized constituent, Truncation is not

possible and null subjects are therefore not expected in these environments.

In order to provide an accurate representation of subject omission with focalized constituents

in the corpus, we examined only predicative utterances which undergo leftward movement

and which require the overt phonological realization of the subjects. Only non-subject

focalized constituent questions were included in the analysis. The data reveal that of the 438

focalized predicates, as detailed in Tables 3 and 4, the subject was always present, except for

two utterances (i.e. 0.5% production), as presented in (56) and (57). These are believed to be

due to errors in performance.

56) A momi Ø niem. (COL 1;11)

FOC mommy Ø name

“It is mommy that she is called.”

57) A kii Ø rait. (ALA 2;02)

FOC key Ø write

“It is a key that I am writing.”

This therefore suggests that there are no grammatical options allowing the focalized particle a

to co-occur with utterances bearing null subjects. This is expected in keeping with the

discussion in the previous chapter. Also recall that the left-ward movement operations

attested in focalized utterances plausibly involve similar movement operations as that attested

in the movement of wh-elements in wh-questions. Clearly, in producing these utterances the


219

CP layer is needed. Being that the subject layer is lower; the subject must also be spelled-out,

thus not permitting null subjects.

8.3.2 Omission of Overt Focal Marker a

Close examination of the data reveals that children produced some utterances which are

presumably focalized, involving left-peripheral movement, but for which the overt focal

marker a is omitted. Initially, these constructions appeared to resemble topicalized

constituents involving overt left-ward movement of the predicate as evidenced by the OSV

order as demonstrated in (58) – (60), but upon closer examination they proved to be cases of

left-peripheral focalization where the overt focus marker is not phonologically realized.

58) Ø Kyahri mi a kaal. (ALA 2;09)

FOC Kyahri 1SG PROG call

“It is Kyahri that I am calling.”

59) Ø de-so im liv. (RJU 2;10)

FOC LOC 3SG live

“It is there that he lives.”

60) Ø ais-kriim mi bai. (COL 2;05)

FOC ice-cream 1SG buy

“It is ice-cream that I bought.”


220

Additional examples presented in (61) – (63) demonstrate that the omission of the focus

marker is also attested in phrases involving prepositional and adjectival fronting.

61) Ø oota Kraasruod mi a_go get i. (KEM 3;01)

FOC out-at Crossroads 1SG PROS get 3SG

“It is out by Crossroads that I am going to get it.”

62) Ø pan yaso mi a jraa wan big ous. (TYA 3;02)

FOC on LOC 1SG PROG draw DET big house

“It is on here that I am drawing a big house.”

63) Ø ogli mi a kuom it. (SHU 3;00)

FOC ugly 1SG PROG comb it

“It is ugly that I am combing it.”

In order to correctly identify and qualify their status as being focalized constituents, native

speakers’ judgments were employed. Two main assessment criteria were used: 1) Restriction

posed on the inclusion of a resumptive pronoun at the end of the focalized utterance in

contrast to the use of the resumptive pronoun with topicalized arguments; and 2) focus is

roughly connected to new information as opposed to the characteristic of topics being

associated with old information that is available and salient in the previous discourse

(Durrleman-Tame, 2008).


221

Table 5 below demonstrates the attestation of predicative focalized utterances in the corpus

without the overt phonological realization of the focus marker. Note that non-subject wh-

questions were not included in this analysis as the focus a is optional with wh-interrogatives.

AGE Total OBJ ADJ PP ADV

2;0.0 1 1 0 0 0

2;0.5 0 0 0 0 0

2;1.0 0 0 0 0 0

2;1.5 0 0 0 0 0

2;2.0 0 0 0 0 0

2;2.5 0 0 0 0 0

2;3.0 0 0 0 0 0

2;3.5 2 2 0 0 0

2;4.0 0 0 0 0 0

2;4.5 1 0 0 1 0

2;5.0 1 1 0 0 0

2;5.5 5 4 0 1 0

2;6.0 1 1 0 0 0

2;6.5 2 2 0 0 0

2;7.0 2 2 0 0 0

2;7.5 5 5 0 0 0

2;8.0 3 3 0 0 0

2;8.5 2 1 0 0 1

2;9.0 4 3 0 0 1

2;9.5 10 6 0 4 0

2;10.0 5 5 0 0 0

2;10.5 2 1 0 1 0

2;11.0 9 7 0 1 1

2;11.5 7 7 0 0 0

3;0.0 9 5 1 3 0

3;0.5 0 0 0 0 0

3;1.0 2 1 0 1 0

3;1.5 14 11 0 2 1

3;2.0 2 1 0 1 0

3;2.5 17 12 0 5 0

3;3.0 4 2 0 2 0

3;3.5 3 2 0 1 0

3;4.0 0 0 0 0 0

3;4.5 1 1 0 0 0

Total 114 86 1 23 4

Table 5: Production of Predicative Focalized Utterances without overt focus marker


222

The examination of the data reveals 114 cases of focalized fronted non-subject constituents

where the focus marker a was not phonetically realized. This represents 60% of the 191

utterances produced demonstrating focalization without the pronunciation of the overt marker.

As shown above, the omission of the focus marker is most dominant in environments where

an object DP undergoes movement to the left periphery of the clausal structure as opposed to

the movement of adjectives prepositions or adverbs. When compared to the overall

predicative focal structures in the corpus, the omission of the focus marker is quite robust

(20.7%). This can be interpreted as another manisfestation of the Privilege of the Root as was

observed in the previous chapter.

8.3.3 Focalized Fronted Constituents with omitted Focus Marker and Null Subjects

Recall that in Section 8.3.1 above we saw that null subjects are not permitted in predicative

utterances with overt pronunciation of the focus marker, as it was attested in a mere 0.5% of

the utterances (i.e. 2 utterances to be exact). This is totally expected in keeping with the

Truncation hypothesis and the revised conception to the ‘Privilege of the Root’ as detailed in

the previous chapter. It is of importance then to examine the dimension of null subjects in

similar predicative utterances demonstrating leftward movement, but where the focalized

marker is omitted.

The data reveals 6% null subjects (i.e. 7 of 114 utterances) in utterances where the focus

marker is omitted. These utterances are presented in (64) – (70):

64) Ø dis kyar Ø av. (COL 2;07)

FOC LOC car Ø have


223

“It is this car that I have.”

65) Ø mai yaad Ø de. (RJU 2 ;09)

FOC POSS.1SG yard Ø LOC

“It is my home that I am at.”

66) Ø op a mi momi Ø liv. (SHU 2;06)

FOC up at 1SG mommy Ø live

“It is up by my mommy that I live.”

67) Ø bog Ø bi. (TYA 2;06)

FOC bug Ø be

“A bug is what it is.”

68) Ø at waata Ø bi. (TYA 2;11)

FOC hot water Ø be

“Hot water is what it is.”

69) Ø Toya Ø de. (TYA 2;08)

FOC Toya Ø LOC

“It is Toya’s (house) that I was.”

70) Ø op a Toya Ø de. (TYA 2;09)

FOC up at Toya Ø LOC

“It is up by Toya’s (house) that she is.”


224

As detailed above, there are only seven utterances in which the subject is omitted following a

fronted predicative focalized utterance for which the focus marker is unpronounced. Closer

examination of the data revealed that these were predominantly attested in one informant’s

utterances: TYA produced 4 of the 7 deviant utterances. SHU, RJU and COL produced one

each, and it was never attested in the speech of ALA and KEM. I would therefore propose that

these utterances are primarily performance errors and do not reflect a grammatical option

allowed by UG.

In line with the Truncation hypothesis and the revised spell-out mechanism presented in

Chapter 7, the overt focus marker a is located in the left periphery, but importantly at a higher

layer than the focalized constituent. The child grammar is therefore assumed to allow the

maximization of clause-initial ellipsis, thereby not spelling-out the left-most elements. In

keeping with “the Privilege of the Root”, the omission of the focus marker in focalized

utterances is therefore permitted until the UG-consistent parametric option of Truncation is no

longer available, around the end of the third year of life. Despite the option to omit the focus

marker, the child needs to build the clausal structure all the way to the CP layer in order to

host the focalized constituent. Being that the subject layer is lower, the subject must be

spelled-out as Truncation is at a point higher than SubjP, thus null subjects are not permitted.

This is contrary to the case of null wh, where Truncation is at a point lower than SubjP,

thereby allowing root null subjects. Under this analysis the data is in line with the findings of

the previous chapter and for overt focalization.

But one questions why TYA produced these deviant utterances. The raw numbers are small,

but then this represents 14% (4 of 28 sentences) of null subjects with fronted predicate for

which the focus marker was omitted; while subject omission was never attested in her


225

utterances where the focus marker was overtly realized. This seems to be problematic,

however based on the limited quantitative data we are unable to delve further into resolving

this issue. Further experimental research would be needed to ascertain if these were mere

performance errors or if it was a genuine grammatical feature in her production.

We will now examine the acquisition of topicalized utterances in children acquiring JC.

8.4 ACQUISITION OF TOPICALIZED CONSTITUENTS IN JC

Topicalization is not overtly marked in JC, but argument topicalization may be optionally

associated with the locative ‘de’, generated to the right of the topicalized constituent.

Additionally, as attested in Italian (Rizzi, 1997), JC argument topicalization entails the

presence of a pronominal element in the comment to refer back to the topicalized constituent.

Adjunct topicalization however does not utilize these options (Durrleman-Tame, 2008).

Our corpus reveals that the productive use of topics begins early, around 23 months. Only

main clause topicalization is produced however, even though the option of embedded clause

topicalization is available in the target language. The data provide evidence for topicalization

of subjects, objects and other adjuncts as shown in (71) – (73) respectively:

71) Di bwai iih mizarebl. (COL 2;07)

DET boy 3SG miserable

“The boy is miserable.”

72) Lait, si i lait ya ! (SHU 2;06)


226

light see DET light LOC

“The light, here is it!”

73) Evri taim dem a dorti-op i kyar. (RJU 2;10)

every time 3PL PROG dirty-up DET car

“Everytime they are dirtying up the car.”

The non-literal use of the locative ‘de’ as is optionally used in the target language to signal an

association of a fronted topicalized object with old information in the discourse is not attested

in the corpus. This option may not be yet available to the children examined. Also the use of a

resumptive pronoun in the comment appears to be optional for the children as it is sometimes

replaced by the topicalized element itself (as in 72).

Even though subjects can be topicalized, their topicalization may be quite ambiguous. It is

therefore difficult to recognize a topicalized subject without the use of any overt marker or

evidence from movement. What I have calculated for topicalized subjects are only the

utterances which signal topicalization via means of repetition or use of a resumptive pronoun

(as in 71) and utterances placing special emphasis on the subject by use of demonstratives and

locatives (as in 74). Therefore, the data presented for topicalized subjects are to be interpreted

with some caution.

74) Da guot ya luk ogli. (ALA 3;00)

DEM goat LOC look ugly

“This goat looks ugly.”

Table 6 outlines the distribution of topicalization in the corpus. We see where objects are

primarily topicalized (76 utterances) followed by subjects (59 utterances). Topicalization of

adjuncts is rarely attested (14 utterances).


227

AGE (mths) TOTAL SUBJECTS OBJECTS ADJUNCTS

1;8.0 1 0 1 0

1;8.5 0 0 0 0

1;9.0 0 0 0 0

1;9.5 0 0 0 0

1;10.0 0 0 0 0

1;10.5 0 0 0 0

1;11.0 0 0 0 0

1;11.5 1 1 0 0

2;0.0 3 2 0 1

2;0.5 2 0 2 0

2;1.0 2 0 2 0

2;1.5 6 0 6 0

2;2.0 3 2 1 0

2;2.5 1 1 0 0

2;3.0 4 3 1 0

2;3.5 4 0 4 0

2;4.0 1 0 1 0

2;4.5 3 0 2 1

2;5.0 3 1 2 0

2;5.5 9 2 5 2

2;6.0 4 2 2 0

2;6.5 8 6 2 0

2;7.0 7 5 2 0

2;7.5 12 9 3 0

2;8.0 7 3 4 0

2;8.5 6 4 2 0

2;9.0 3 1 2 0

2;9.5 7 2 5 0

2;10.0 8 1 5 2

2;10.5 3 0 2 1

2;11.0 10 5 3 2

2;11.5 6 1 4 1

3;0.0 6 3 2 1

3;0.5 4 2 2 0

3;1.0 2 0 0 2

3;1.5 2 0 1 1

3;2.0 3 0 3 0

3;2.5 4 1 3 0

3;3.0 0 0 0 0

3;3.5 3 1 2 0

3;4.0 0 0 0 0

3;4.5 1 1 0 0

TOTAL 149 59 76 14

Table 6: Production of Topicalized Utterances


228

As revealed in Table 7, topicalization, like focalization, is productive during the 2-words

combination stage. This finding shows that both phenomena demonstrating left-peripheral

movement are evident early in the production of children acquiring JC. We see however

where the focalization of objects and other predicates is not attested until the 3-words

combination stage, topicalization of objects and other predicates are attested at the very initial

2-words combination stage. Predicative topicalization therefore seems to be more easily

acquired than predicative focalization by children learning JC.

MLU TOTAL SUBJECTS OBJECTS ADJUNCTS

< 1.59 0 0 0 0

1.5 – 2.49 14 7 6 1

2.5 - 3.49 52 23 26 3

3.5 - 4.49 40 15 23 2

4.5 > 43 14 21 8

Table 7: Production of topicalized utterances with MLU

8.4.1 Target inconsistencies in early Topicalized Utterances

Closer examination of the topicalized utterances revealed that the data consist of target

inconsistencies in the form of omissions of subjects. In order to examine such target

inconsistencies, we looked at the 90 topicalized objects and adjuncts which had undergone

movement to the left periphery of the clause. In keeping with our discussion on the UG option

of Truncation and the ‘Privilege of the Root’, we would expect that subject omission would

not be permitted in topicalized utterances. This is based on the idea that the left periphery

contained an overt element, the topicalized object, which is located higher in the clausal

structure than the subject. Contrary to our expectations however, we saw a vast number (21

utterances or 23%) of null subjects in these environments. This called for a more detailed


229

analysis of left-peripheral topicalization in JC. Presented in (75) – (95) are the utterances to be

examined.

75) Man Ø kik i. (COL 1;08)

man Ø kicked 3SG

“The man, I kicked him.”

76) Aad Ø push iin i. (ALA 2;00)

hard Ø push in 3SG

“Hard, I pushed it in.”

77) Kyahri Ø bait Ø. (ALA 2;00)

Kyahri Ø bait Ø

“Kyahri, he bit him.”

78) Chiiz-chiks Ø dash i we. (COL 2;01)

chiiz-chiks Ø dash 3SG away

“The cheeze-trix, I threw it away.”

79) Kidi Ø fiid im. (ALA 2;01)

kiddy Ø feed 3SG

“Kiddy, I fed him.”

80) Glaas Ø put aan Ø. (RJU 2;01)

Glass Ø put on Ø


230

“The pair of glasses, I put it on.”

81) Baisikl Ø si i baisikl. (RJU 2;01)

bicycle Ø see DET bicycle

“The bicycle, I see the bicycle.”

82) Kaada Ø pres kaada? (RJU 2;01)

recorder Ø press recorder

“The recorder, Am I to press the recorder?”

83) AJ baik Ø a raid i baik. (COL 2;02)

AJ bike Ø PROG ride DET bike

“AJ’s bike, he is riding the bike.”

84) Van Ø wahn sopm in de. (KEM 2;04)

van Ø want something in LOC

“The van, I want something in there.”

85) Buk Ø riid buk. (RJU 2;05)

book Ø read book

“The book, I read the book.”

86) Baisikl Ø jraiv Ø. (KEM 2;05)

bicycle Ø drive Ø

“The bicycle, I am going to ride it.”


231

87) Gyas Ø gi it gyas. (KEM 2;05)

gas Ø give 3SG gas

“Gas, I gave it gas.”

88) Popi Ø av Ø. (TYA 2;06)

poppy Ø have Ø

“A poppy, I have a poppy.”

89) Baik Ø go fi i baik ya now. (KEM 2;07)

bike Ø go for DET bike LOC now

“The bike, I am going for the bike right now.”

90) Mashiin Ø yuuz i mashiin mi ed. (KEM 2;07)

machine Ø use 3SG machine 1SG head

“Machine, I use it to machine my head.”

91) Sombadi Ø kaal sombadi. (RJU 2 ;07)

somebody Ø call somebody

“Someone, I am going to call someone.”

92) AJ Ø gaan im breda. (COL 2;08)

AJ Ø gone 3SG brother

“AJ, she went to her brother.”


232

93) Dat Ø tek op dat. (COL 2;08)

DEM Ø take up DEM

“That, he took up that.”

94) Waal Ø krash waal. (TYA 2;08)

wall Ø crash wall

“The wall, he crashed the wall.”

95) Rait de-so nou Ø mek i stie rait de-so. (RJU 3 ;00)

right LOC now Ø make 3SG stay right LOC

“Right there now, I am going to let it stay right there.”

After closer analysis of the null subjects following topicalized constituents, it appears as if

topicalization in JC is amenable to an analysis whereby the phrase must be seen as two

separate structural pieces. The topic is plausibly a separate intonational phrase, having distinct

intonation from the comment it precedes. For JC, the boundary of a non-final intonational

phrase is cued by a pause and a pitch reset to a higher Fundamental frequency in the

immediately following intonational phrase (Gooden, 2014). This seems to be the case for the

examples above demonstrating left-peripheral topicalization, as signaled by the pause32

. This

is contrary to focalization as discussed in the previous sections, where there is no perception

of a phrase break after the focalized constituent, such that it may be argued that the post-focal

accent is in a different phrase. Moreover in JC, there is a clear intonational contrast between

the pitch accent on neutral broad focus declarative sentences (as in the comment of the clause)

and the focal accent found on the emphatic clause/topic in the intonational phrase. The latter

32 The actual prosodic prominence associated with topics was not measured.


233

is observed in pragmatic contexts requiring confirmation of old information (typically

associated with topics) and also just before a high Fundamental frequency, which might be

analyzed as a pitch reset marking the start of the next phrase (see Gooden 2014). These

factors would argue in favor of the topicalized constituent being an independent structural

entity to the comment it precedes. This analysis is similar to that of Italian in that there is a

sharp prosodic distinction between topic-comment articulation where the post-topic material

is viewed as a separate entity, as opposed to focus-presupposition (see Bocci 2007; Bianchi,

Bocci and Cruschina 2013).

If this analysis of topics being a separate structural entity is plausible, then arguably

topicalization in JC could be seen somewhat like left dislocation in English. The dislocated

topic is non-syntactic, in the sense that it does not participate in the subject-predicate structure

of the clause. Crucially however, the topicalized object must be related to the remaining part

of the following clause (i.e. it must agree in number and gender), thereby requiring an

obligatory resumptive element, or a repetition of the topic itself. In quite a few of the

examples, the topic is resumed by a copy (as opposed to a resumptive pronoun) in the clause,

which suggests that it is not ‘hanging’ but directly linked to an element inside the comment.

Being that there are no requirements on the topic to necessarily bind a resumptive pronoun,

arguably, the topic could be CP-external and does not need to fill Spec CP. The topic could be

simply prefixed to an independent root CP, subject only to the requirement that the CP must

be about the dislocated topic (see Aissen 1992 for similar discussion with respect to Mayan

languages). We therefore tentatively propose that these dislocated topics are essentially

external to and separate from the main clausal structure, hence, arguably, not interfering with

the Privilege of the Root as discussed in the previous chapter. Importantly however left-


234

dislocated topics which are seemingly anchored by a locative adverb and which must bind a

resumptive pronoun (along the lines of Durrleman-Tame 2008) must be CP-internal.

One major question stands out: why should topics in the children’s production behave so

differently from other left-peripheral constructions such as focus and wh-movement with

respect to the null subject option? For us to gain a better understanding of this phenomenon

further research, possibly including experimental analysis of the acquisition of topicalization

in JC is required.

8.5. CARTOGRAPHIC ORDERING OF THE LEFT PERIPHERY IN JC

In this chapter I discuss material which appears to be located in the left periphery of the

clausal structure. We will now look at the cases where more than one such element co-occurs

in the CP. Based on the co-occurrence of left-peripheral elements in the clausal structure, a

single projection CP cannot suffice, and as such I will adopt Rizzi’s (1997) articulate structure

of the CP also known as the Split CP Hypothesis. Rizzi (1997) upholds that the

complementizer system consists of an array of projections articulated along the lines of (96)

below:

96) ForceP > TopP > FocP > TopP > FinP > IP

It is of interest then to see whether children acquiring JC respect this clausal hierarchy. But

before we venture into such an analysis, it is inevitable that we qualify what is hosted in each

position in JC. We have already examined Topic and Focus in much detail and as such we

will look briefly at Force and Fin. In keeping with Durrleman-Tame (2008) and Bailey

(1966), when the word se is used in certain contexts, for example when it selects sentential


235

complements, it seems to take on the role of a complementizer. This is evident in the

children’s utterances:

97) Mek mi tel dadi se mi wahn sinak. (ALA 2;08)

Make 1SG tell daddy CMP 1SG want snack

“Let me tell daddy that I want snacks.”

98) Yu si se it a blingk faas. (SHU 3 ;01)

2SG see CMP 3SG PROG blink fast

“You see that it is blinking fast.”

By situating se in Force33

, it followed from the structure in Rizzi (1997) that in embedded

contexts, both syntactically focalized and topicalized elements have to follow se (Durrleman-

Tame 2008). Though such complex utterances are rarely attested in the data, the following

demonstrates evidence of this in early JC:

99) Gad nuo [FORCE se [FOC a Manski dat.]] (RJU 2 ;04)

God know CMP FOC Manski DEM

“God knows that that is Manski.”

An element which may sit in Fin is fi34

(Durrleman-Tame, 2008). This element must select the

subject of a non-finite clause and therefore cannot be followed by a topic. An example of the

overt realization of Fin in the corpus is:

33 Durrleman-Tame (2008) argues that se is best analyzed as a lexical verb in an SVC rather than an overt complementizer. She suggested that Force in JC is void of overt material. 34

Parallel to Force, Durrleman-Tame (ibid) posits that morphologically filling Fin is not the preferred option in basilectal JC.


236

100) [FIN Fi [IP dadi kom biit mi.]] (ALA 2;09)

FIN daddy come beat 1SG

“For daddy to beat me.”

Once again these elements are rare in the dataset and as such I will attempt to provide only

marginal evidence for the attested structure. The following examples demonstrate the ordering

possibilities which are found in the children’s production:

101) [TOP Mi [WH we mi du wid di kii?]] (KEM 3;02)

1SG where 1SG do with DET key

“I, what have I done with the key?”

102) [TOP Da kola ya [FOC a yuu a_go get i.]] (TYA 3;01)

DEM colour LOC FOC 2SG PROS get 3SG

“This colour, you are going to get it.”

As shown in (101), topic is seen to precede Wh. The reverse order would yield

ungrammaticality and was never attested in the corpus. This observation also holds for (102)

where focussed constituents may follow topics but not the reverse. Note that in JC the

focussed constituent and the wh-word is in complementary distribution as both plausibly

targets the same position (Durrleman-Tame 2008). No evidence of the co-occurence of a

focalized constituent and a wh-element was detected in the data. The relative order as

demonstrated in (99) through to (102) is as follows:

103) a. Force > Foc

b. Fin > IP


237

c. Top > Wh

d. Top > Foc

The utterances considered here therefore gives direct evidence for the structure in (104):

104) ForceP > TopP > FocP ~ Wh > FinP> IP

This structure is compatible with Rizzi’s (1997) Split CP Hypothesis and demonstrates that

from an early age children acquiring JC are knowledgeable of the rules governing the

ordering of constituents in the left periphery. The attested sequence outlined above confirms

that children acquiring JC do not entertain the possibility of target-inconsistent orders in their

development of the CP. Recall that we have also seen in Chapter 6, that children are also

knowledgeable of the relative order in combination of IP elements. This target-consistent

hierarchical development of structures attested at different clausal levels calls for a full

competence approach to the development of the cartographic sequence.

8.6 DISCUSSION AND CONCLUSION

A great deal of development is demonstrated to occur quite early in the grammar of children

acquiring JC. Though wh-interrogation, focalization and topicalization are all seemingly

complex phenomena analyzed as involving movement to the left periphery (Durrleman-Tame

2008; Durrleman & Shlonsky 2015), they are however evident quite early in development.

Wh-questions are produced during the single word stage where MLU is under 1.49 in the

data; topicalized predicative utterances are acquired just a bit later at the 2-word combination

stage; and Predicative focalization of non-wh-constituents becomes productive during the 3-

words combination stage.


238

The data provides evidence that focalized elements are unable to co-occur with null subjects,

whether the focus marker is overt or phonologically null, thereby providing support for the

Truncation hypothesis and the revised “Privilege of the Root”. The co-occurrence of

topicalized left dislocated elements and null subjects lead us to propose that in early JC,

where there are no requirements on the topic to necessarily bind a resumptive pronoun, the

topic does not need to fill Spec CP. The topic is possibly a separate structural entity which

feasibly is external to the clause and not interfering with the “Privilege of the Root.” Where

the requirement is upheld however, topicalization must be CP-internal, as seen for focus and

wh-movement. The relative ordering of the elements within the CP suggests that children

acquiring JC are aware of the cartographic hierarchy of the target language and use the

various projections in the correct order. We therefore conclude that wh-interrogation,

focalization and topicalization are all complex phenomena, being acquired seemingly

effortlessly and early in the development of JC.

However, at this point many questions still remain: Why should topics in the children’s

production behave so differently from other left-peripheral constructions with respect to the

null subject option? Why should one informant produce null subjects with focalized fronted

predicative utterances when the focus marker is omitted? These questions call for further

research of the early acquisition of the left-periphery in JC, possibly, incorporating

experimental research. Additionally, the intonation patterns associated with these utterances

are doubtlessly a promising venue for further investigation of these issues.

Chapter 9: The Acquisition of Typical Creole Features

239

CHAPTER 9

THE ACQUISITION OF TYPICAL CREOLE FEATURES

9.1 INTRODUCTION

The goal of this chapter is to provide a systematic analysis of the acquisition of some

constructions described as being typically creole. As a starting point, Bickerton (1981 &

1999) claimed that there are some features which are shared by a wide range of creole

languages that show some striking similarities. These features are believed to be contained in

the list of default universal grammaticizations as argued for in his Language Bioprogram

Hypothesis. According to Bickerton (1981), some of these features which are common to

creole languages are the same features that children acquiring non-creole languages produce

effortlessly, not from characteristics of the input but rather from the functioning of the innate

bioprogram. Bickerton (1999) further states that children in acquisition would not necessarily

enforce the default-list distinctions where other types of allowable distinctions are readily

available in target language input. However where the grammaticalized distinctions of the

target language conform to the default distinctions, the child will acquire those distinctions

early, rapidly and without error.

In light of the enormous amount of work done since Bickerton’s basic ideas of typical creole

features were formulated, at this point such features can hardly be considered characteristic of

solely creole languages. I will discuss some of these alleged features in order to ascertain

whether they are acquired differently in JC as opposed to non-creole languages. I will pay

special attention to the production of negation, verb serialization, inclusive/plural marking,

pronouns and reflexives, copulas and determiners in early language development.


240

The chapter is organized as follows. Section 9.2 examines negation, looking at whether the

grammar of children acquiring JC is characterized by the use of target-consistent anaphoric

negation, non-anaphoric sentential and constituent negation, and negative concords. Section

9.3 deals with the acquisition of verb serialization, examining its attestation in the corpus, and

the combinations produced in accordance with the target language. Pluralization/Inclusiveness

is the subject of Section 9.4. Here children’s early production is mapped in order to determine

its similarities with the target grammar. In section 9.5 pronouns and reflexives are examined

in order to ascertain whether children demonstrate knowledge of binding principles. The use

of copulas and determiners in the production of children acquiring JC are studied in Sections

9.6 and 9.7 respectively. The main conclusions are presented in 9.8.

9.2 NEGATION

Negation in JC is predominantly expressed by the preverbal negative element no, however

there are other negators dohn ‘do not’, nat ‘not’, and neva ‘never’. As demonstrated below,

children acquiring JC are aware of the different methods of expressing negation:

1) Mi no fried a krokodail. (TYA 2;11)

1SG NEG afraid of crocodile

“I am not afraid of crocodiles.”

2) Mi dohn sii im. (RJU 2;03)

1SG NEG see 3SG

“I do not see him.”


241

3) I nat workin at’aal. (ALA 2;11)

3SG NEG working at all

“It is not working at all.”

4) Yu neva nuo? (SHU 3;00)

2SG NEG know

“Didn’t you know?”

Additionally negated Tense, Modal and Aspectual (TMA) utterances are also evident in the

children’s utterances:

5) I naa dash we. (COL 2;06)

3SG NEG~PROG dash away

“It is not spilling.”

6) Yo naa_go jrap aaf. (ALA 2;08)

2SG NEG~PROS drop off

“You are not going to fall off.”

7) A no don kuul i. (SHU 3;00)

1SG NEG COMP cool 3SG

“I haven’t finished cooling it/ cooled it completely.”

8) I kyaahn staat. (KEM 2;07)


242

3SG MOD~NEG start

“It cannot start.”

9) A shudn mek im gu we. (ALA 2;08)

1SG MOD~NEG make 3SG go away

“I shouldn’t let him go away.”

10) Mi no ben go a noo skuul. (KEM 3;00)

1SG NEG PAST go at NEG school

“I didn’t go to school.”

11) Wai shi neva did stie (wid) mi. (SHU 2;09)

why 3SG NEG PAST stay with 1SG

“Why didn’t she stay with me?”

12) Mi no fi go in de. (ALA 2;09)

1SG NEG INF go in LOC

“I am not to go in there.”

In these sentences, the negative element is target-consistently placed before the preverbal

TMA markers. For utterances expressing progressive or prospective aspect, there is

coalescence of the negative particle no with the progressive a and prospective a_go yielding

naa and naa_go respectively as seen in examples (5) and (6). For utterances expressing

modality, the negative particle no does not need to co-occur with modals as generally modals


243

have a negative form35

. For example, as seen in (8) and (9) above negative kyaahn and shudn

are differentiated from positive kyahn and shuda respectively. With regards to tense, the

negator is placed before the preverbal tense marker as in no ben and neva did as seen in (10)

and (11) respectively. Also, in cases where the verb is in the infinitive form, the negator is

placed before the infinitival marker as seen in example (12). The examples so far all

demonstrated sentential negation. However, constituent negation, another option available in

the target language, is also attested in the children’s production:

13) A no dadi chok dat. (ALA2;07)

FOC NEG daddy truck DEM

“That is not daddy’s truck.”

14) A nat mii tier i op. (COL 2;07)

FOC NEG 1SG tear 3SG up

“It was not me who tore it up.”

15) A no Brianna a raid i wumba. (KEM 3;00)

FOC NEG Brianna PROG ride DET bike

“It is not Brianna who is riding the bike.”

16) A nat riid mi a riid. (RJU 2;07)

FOC NEG read 1SG PROG read

“It is not read that I am doing.”

35 Where modals are seen to co-occur with the negative element, two constructions are possible: one in which the negator occurs before (as in mi no kuda kom wid yu “couldn’t I have come along with you?”) and one in which the negator occurs after (as in mi kuda no kom wid yu “I could have not come along with you (but I did)”). Note the difference in interpretation when the negative modal is used: mi kudn kom wid yu “I could not come with you.”


244

17) A no chok, a kyar. (SHU 3;00)

FOC NEG truck FOC car

“It is not a truck, it is a car.”

18) A nat so yu put aan i. (TYA 3;02)

FOC NEG so 2SG put on 3SG

“It is not like this that you put it on.”

In constituent negation, clearly the negative element follows the focus marker, but

importantly precedes the negated constituent. The data reflect no errors in these utterances,

showing that JC children have no difficulties in acquiring sentential and constituent negation.

Children have demonstrated knowledge that the negative element should be placed to the left

of its complements, hence the correct placement of negation.

Whereas children acquiring JC demonstrate no difficulties in the production of constituent

negation, the contrary is reported for children acquiring Mauritian Creole (MC). Based on a

pilot study conducted in 1987 (as reported in Adone, 1994), children up to 3;03 years had

difficulties repeating sentences with constituent negation. This is due to the seemingly

complicated pattern for producing constituent negation in MC, which employs the overt

complementizer ki, as exemplified in (19) below:

19) Pa papa ki ti tay gazoñ (MC: Adone, 1994)

NEG father CMP36

TNS cut grass

36

Original code for complementizer COMP changed to CMP so as not to confuse data with the code COMP as used in this work for the completive aspect.


245

“It is not father who cut the grass.”

According to Adone (1994), it is not the negation which is difficult to acquire in constituent

negative utterances, but the movement rule together with an overt CMP element which

complicates matters. JC need not employ an overt CMP element with constituent negation and

as such children produce these utterances quite early and without errors.

Also noted in the corpora are target-consistent utterances with a negative element followed by

a sentence nucleus. Some examples are presented in (20) to (25).

20) Noo biebi in mi lap. (ALA 2;04)

NEG baby in 1SG lap

“No, the baby is in my lap.”

21) Noo mi naa jrap. (COL 2;06)

NEG 1SG NEG~PROG drop

“No, I am not falling.”

22) Noo a fi mi i fa. (KEM 2;06)

NEG FOC POSS 1SG 3SG for

“No, it is mine.”

23) Noo a mai baks. (RJU 2;07)

NEG FOC POSS.1SG box

“No, it is my box.”


246

24) Noo im gaan a Merika. (SHU 2;05)

NEG 3SG gone to America

“No, he went to America.”

25) Noo a likl kyat. (TYA 2;09)

NEG FOC little cat

“No, it is a little cat.”

In these utterances, the overt subjects are positioned to the right of the negative element. They

are target-consistent anaphoric negation in which the negative element is in a sentence

peripheral position, and acts to negate a prior utterance. Unlike for children acquiring English

where similar utterances are used ungrammatically in a non-anaphoric manner (as reported by

Klima & Bellugi 1966; Gilkerson, Hyams & Curtiss 2003), children acquiring JC tend to be

target-consistent.

A look at Table 1 shows that children already have sentential/constituent no during the single

word stage. Closer examination reveals that all 10 utterances during this stage were produced

by one informant TYA. We see at the 2 words combination stage a vast increase in the use of

dohn, being the primary negative element at this stage. Closer examination reveals that 96%

of its production was solely by ALA as RJU produced only 3 utterances and COL produced 1.

It was unattested in the speech of the other informants. An examination of ALA’s files shows

where dohn was not being used productively during this stage as 97% of the times it was used

with just a single verb noo/nuo ‘know’. At the 3 words combination stage we see a massive

increase in the use of no by all informants. All informants now use dohn but it is still


247

predominantly employed (52%) in the speech of ALA. The use of naa and kyaahn do not

reach steady frequencies until later in the corpus when children are producing more complex

multiword utterances. Neva is rarely attested in the speech of the informants.

Mlu No Dohn Naa Kyaahn Nat Neva Anaphor noo

< 1.49 5 0 0 0 0 0 5

1.5 - 2.49 30 112 5 1 2 0 48

2.5 - 3.49 191 121 78 43 14 2 78

3.5 - 4.49 310 115 206 181 32 10 102

4.5 > 731 215 446 228 79 35 307

TOTAL 1267 563 735 453 127 47 540

Table 1: Negation 37

Note that there is not a single example of wrong placement of negation in the data. Also, there

is no stage in which the negative element occurs only in sentence-initial position. This

provides cross-linguistic support for Clahsen (1988) and Adone (1994), among others, who

claim that both non-anaphoric and anaphoric negation occur at the same time, contra Wode’s

(1977) proposal that anaphoric negation precedes non-anaphoric negation. As a sentential

negator, the negative element is always placed to the left of the verb or VP. This therefore

shows that no mistakes are made in the acquisition of NEG, since from the beginning,

children are aware of its correct placement. In this respect, the negation pattern of children is

target-consistent from the start.

During the early stages however, we have noted utterances in which the obligatory subject is

dropped in both anaphoric and non-anaphoric sentences as in (26) and (27) respectively. Even

in these sentences, there is no evidence in support of a hypothesis suggesting that children

acquiring JC make mistakes in terms of the placement of negation.

37

Only sentence and constituent negation were included in this count. Single-word non-affirmative negation was not included.


248

26) Noo Ø tek im we. (KEM 2;07)

NEG Ø take 3SG away

“No, (he) took him away.”

27) Ø no av i buk. (ALA 2;08)

Ø NEG have DET book

“I do not have the book.”

Following Pollock (1989), Deprez and Pierce (1993), among others, we affirm that non-

anaphoric negation is situated in the same position in both child and adult grammar, forming

part of the sentence’s inflectional projections. On this view, it is generated below IP and

above VP, as is the case for Mauritian Creole child grammar around 28 months of age

(Adone, 1994), as represented in the simplistic diagram below:

This shows that children acquiring JC raise the subject NP above negation, thereby producing

target-consistent negation to the right of the subject (whether overtly pronounced or not). The

absence of utterances where a non-anaphoric negative element occurs before the subject NP

IP

NegP

VP

Spec

NEG

NP V


249

provides evidence that non-anaphoric negation is situated in an IP-internal position. On the

other hand, anaphoric negation is arguably situated in a sentence-peripheral position, rather

than sentence-internally. Both types of negation may co-occur in the production of the

children, with or without an overt subject, as demonstrated in (28) and (29) respectively.

28) Noo mi no wiek i biebi. (SHU 2;05)

NEG 1SG NEG wake DET baby

“No, I didn’t wake the baby.”

29) Noo Ø no wahnt i. (COL 2;08)

NEG Ø NEG want 3SG.

“No, (I) don’t want it.”

This seems to be a logical departure point for us to now venture in the case of multiple

negation.

9.2.1 Multiple Negation

Another development can be observed in the children’s corpora - the emergence of multiple

negation:

30) Nobadi naa_go si yo. (ALA 2;08)

Nobody NEG~PROS see 2SG

“Nobody is going to see you.”


250

31) Im naa mek op no naiz. (COL 2;07)

3SG NEG-PROG make up NEG noise

“He is not making any noise.”

32) Mi no av no skuul bag. (KEM 2;08)

1SG NEG have NEG school bag

“I don’t have any school bag.”

33) Yo naa get non. (RJU 2;09)

2SG NEG~PROG get none

“You are not getting any.”

34) Mi kyaahn put aan no muor. (SHU 2;07)

1SG NEG~MOD put on NEG more

“I cannot put on any more.”

35) Yo neva gi mi no moni. (TYA 3;02)

2SG never give 1SG NEG money

“You never gave me any money.”

As seen above, multiple negation may take the form of the co-occurrence of two (or more)

negative elements, or it can be a negative marker co-occurring with a negative word.

According to Bickerton (1981) this is a feature of creole languages, however it is not

exclusive to creoles. It is a well-known phenomenon in some Romance languages, the Slavic


251

languages and Russian (Syea, 1993). In JC, multiple negatives have been described as

negative concords (Patrick, 2004).

Detailed in Table 2, we see where all informants produced at least one utterance with negative

concord during the initial multiword utterance stage where MLU is between 2.5 and 3.49.

Once again there are no mistakes in the placement of the negative elements, and the data

reflect immediate target consistency.

MLU COL ALA RJU TYA KEM SHU TOTAL

< 1.49 - 0 0 1 - - 1

1.5 – 2.49 0 0 0 0 0 - 0

2.5 – 3.49 3 2 1 2 1 9 18

3.5 – 4.49 22 9 13 4 6 12 66

4.5 > - 38 3 20 102 25 188

TOTAL 25 49 17 27 109 46 273

Table 2: Multiple Negation

Not surprisingly however, the data shows that children start producing negative concord

constructions after sentences with a single negative element. Various studies have shown that,

although ungrammatical in Standard English and thus absent from English-speaking

children’s input, children acquiring English actually produce these structures (Bellugi 1967,

Brown 1973, Thorton & Tesan 2013). Whereas such utterances would be ungrammatical in

children acquiring Standard English, in JC they are target-consistent. The data therefore is in

line with Bickerton’s claim that double negation is one of the typical creole features that

children acquiring creole languages acquire without errors, however being that it is also a

common feature of non-creole languages, it is not clear whether this shows anything relevant

for the Bioprogram Hypothesis.


252

9.2.2 Target inconsistency

Notwithstanding the early production and correct placement of negation in the corpora, we

have noticed some examples of target inconsistency in the children’s utterances. These range

from errors of omission to errors of commission. We have noted 31 utterances where the

obligatory negative marker is omitted as in (36) and (37).

36) Ø Ø want it. (ALA 2;00)

Ø Ø want 3SG

“(I don’t) want it.”

37) Ø Ø plie wid im. (COL 2;03)

Ø Ø play with 3SG

“(I am not) playing with him.”

These omissions are only identifiable based on the context of the discourse in which they are

produced. Note that all but 4 such examples for which the negative elements are omitted, the

subjects are also omitted. The 4 examples with the overt subjects arguably reflect errors in

performance, while elsewhere children are likely to be exercising the option available to them

for truncating the clause. This follows from the idea as discussed previously in Chapters 7 and

8, that there is a lower Truncation in these cases, hence subject omission is possible.

Regarding errors of commission, we note that there are 12 such errors, all of which

demonstrate the use of an incorrect negative particle as in (38) and (39):


253

38) Iih *naa ha noo tiit. (COL 2;10)

3SG NEG~PROG have NEG teeth

“He doesn’t have any teeth.”

39) Mi no waahn *non aki. (TYA 2;11)

1SG NEG want NEG ackee

“I don’t want any ackee.”

In (38) and (39) the correct negator should be no. As they are not robustly attested, these

errors could be attributed to errors in performance.

We will now turn our attention to verb serialization.

9.3 SERIAL VERB CONSTRUCTION

The notion of serial verb construction (SVC) is highly problematic in the literature. According

to Sebba (1987:1) “It is not at all clear that all authors are referring to the same thing when

they speak of “serial verbs”.” According to Adone (2012:53) a SVC is “a chain of verbs

which functions as a single predicate without coordination, subordination, or syntactic

dependency”. She later defined SVCs as “complex predicates containing at least two verbs

within a single clause.” (Adone, 2012:144). The latter definition is also shared by Veenstra

(1996) and will be adopted for the classification of SVCs analyzed in this chapter. 8 major

groups of SVCs are attested in the corpora as exemplified in (40) to (47) below:

40) Directional: A jraiv go skuul. (COL 2;04)


254

1SG drive go school

“I drove to school.”

41) Causative: Yaa_go mek momi ded fi ongri. (ALA 2;11)

2SG~PROS make mommy dead for hungry

“You are going to cause Mommy to die of hunger.”

42) Instrumental: Ø Ø tek yuuz i raid baik. (KEM 2;07)

Ø Ø take use 3SG ride bike

“I am using it for a bike.”

43) Benefactive: Chuo i gi im! (ALA 2;06)

throw 3SG give 3SG

“Throw it to him”

44) Intentional: Jerwayne kom gi mi i bag! (TYA 2;10)

Jerwayne come give 1SG DET bag

“Jerwayne give me the bag!”

45) Resultative: A se i man baks ar bos ar nii. (COL 2;08)

1SG say DET man box 3SG burst 3SG knee

“I said that the man boxed her causing her to cut her knee.”

46) Dative: Mi aa ful wata gi im. (RJU 2;11)

1SG PROS full water give 3SG


255

“I am going to fill (a bucket of) water and give (it) to him.”

47) Say: We yaa taak tu mi se? (SHU 2;09)

What 2SG~PROG talk to 1SG say

“What are you saying to me?”

The data reveal that children start to produce SVCs at an early age, during the initial 2-words

combination stage where MLU is between 1.5 and 2.49. However the overall frequency of

occurrence is low during the entire developmental period studied. Only 180 utterances with

SVCs were found in the corpus as detailed in Table 3:

MLU DIR CAU INS BEN INT RES DAT SAY

< 1.49 0 0 0 0 0 0 0 0

1.5 - 2-49 6 1 0 0 3 0 1 0

2.5 - 3.49 15 0 1 3 13 2 0 0

3.5 - 4.49 10 6 3 3 10 2 2 1

4.5 > 29 6 0 1 48 11 1 2

TOTAL 60 13 4 7 74 15 4 3

Table 3: Serial verb constructions with MLU

The corpus shows the following picture: for the entire period studied, children demonstrate

productive use of mainly the directional and the intentional SVCs. For directional SVCs, the

verbs kom ‘come’ and go ‘go’ are in non-initial positions as in the target language. They are

used to mark direction away or towards a point of reference. They are used in combination

with verbs of movement, as in (48) – (50); however there are no strict restrictions on the type

of verb they combine with in the target language.

48) Kidi ron kom. (KEM 2;02)

kidi run come


256

“The kid ran towards me.”

49) Mi aa haal im go ova de-so. (RJU 2;11)

1SG PROS haul 3SG go over LOC

“I am going to drag him over there.”

50) AJ kyari it kom! (SHU 2;03)

AJ carry 3SG come

“AJ carry it to me!”

For intentional SVCs, the verbs kom ‘come’ and go ‘go’ are used again but now in initial

positions. This structure is in conformity with the target language. They are used to mark

deliberate intent of an action, as in (51) – (53) and are not restricted to co-occur with any

specific verb class.

51) Mi a_go jraiv i van kom krash yo. (KEM, 2;08)

1SG PROS drive DET van come crash 2SG

“I am going to drive the van and (intentionally) crash it in you.”

52) Go tek-op mi slippaz! (RJU 2;00)

Go take-up 1SG slippers

“Go (intentionally) and take up my slippers!”

53) Grandad kom kot out yaso fi mi. (SHU 3;01)

Grand-dad come cut out LOC for 1SG


257

“Grandfather (intentionally) cut here for me.”

The frequent appearance of these V+go/kom and go/kom+V structures in the children’s

utterances confirms their productive use in early child grammar.

During the later multiword utterances where MLU is above 3.5, we note the increase in use of

causative and resultative SVCs. Though the data is scant, it strongly suggests that these

structures are acquired later than the go/kom constructions discussed above. The verb mek

‘make’ is the causative serial in both the children’s utterance and the target language. It is

normally in first position, mek+V, or it can be the middle verb in a three verb construction,

V+mek+V, as shown below:

54) Mi a_go mek worm it dem. (KEM 3;02)

1SG PROS make worm eat 3PL

“I am going to cause worms to eat them.”

55) Plie i mek a daans! (SHU 2;08)

Play 3SG make 1SG dance

“Play it so that I can dance!”

56) Put im dong Adrian mek im ron. (ALA 2;06)

Put 3SG down Adrian make 3SG run

“Put him down Adrian allowing him to run.”


258

For resultative SVCs, the result-denoting verb is more or less fixed in a non-initial position as

seen in (57) to (59). The list of verbs here is unrestricted.

57) Mi a go iin lef yo. (KEM 3;02)

1SG PROG go in leave 2SG

“I am leaving you.”

58) Ø raid i jrap. (COL 2;06)

Ø ride 3SG drop

“I fell (off the bicycle).”

59) Im a tek som graas a iit. (ALA 3;00)

3SG PROG take some grass PROG eat

“He is eating some grass.”

Instrumental, benefactive, dative and serial verbs with ‘say’ are hardly attested in the data. As

illustrated in the examples above, only certain verbs can be combined with others, and in a

specific format, depending on the construction being targeted. The fact that children make no

mistakes in the structure and combination of these serializing verbs means that they conform

to the adult model and demonstrate knowledge of the rules in SVC formation from an early

age.

Closer analysis of the data reveals huge individual variation in the corpus. Table 4 shows

where KEM produced 62 (34%) of the SVCs in the corpus followed by SHU who produced


259

40. ALA and RJU produced 30 and 29 respectively. SVCs were rarely attested in COL’s and

TYA’s utterances (13 and 6 respectively).

INFORMANTS DIR CAU INS BEN INT RES DAT SAY TOTAL

COL 7 2 0 0 2 2 0 0 13

ALA 9 2 1 1 13 2 1 1 30

RJU 12 2 0 2 9 2 2 0 29

TYA 2 1 0 0 3 0 0 0 6

KEM 19 4 3 1 26 8 0 1 62

SHU 11 2 0 3 21 1 1 1 40

Table 4: Individual production of SVCs

At this point, it becomes clear that the presence of SVCs in the data has to be accounted for.

We aim to do so in the subsequent section.

9.3.1 Accounting for Serial Verb Constructions in Creole Acquisition

According to Adone (2012), the ease with which children produce SVCs can be taken as a

strong indication that children’s grammars generate these sentences. This would further seem

to suggest that some serialisation pattern/structure is plausibly a part of UG that generates a

wide range of these structures. Evidence of this serializing tendency in acquisition is

confirmed in studies of non-serialising languages: for example Bowerman (1982) reported

that causative SCVs are found in early child English.

Accounting for SVCs is an area of controversy in the literature. Adone (2012), for example,

proposed that in children’s initial phrase structure configuration, elements are attached in an

adjoined position, rather than a complement position (in line with Roeper and de Villiers

1992, Hoekstra and Jordens 1994, Roeper 1996). Therefore case marking and

subcategorization are defined in lexical rather than structural terms. Frank (1992) also shows


260

that children assign a conjoined structure [NP1 V NP2] [e1 V NP3] to structures such as NP1 V

NP2 V NP3, whereby [e1 V NP3] is treated as an adjunct, showing children’s preference for

subject control of the embedded subject. Of the two principles available in the target grammar

(i.e. complements and adjuncts), the adjunction configuration and subject control together

account best for the early SVCs in children’s grammar. This adjunction configuration is likely

to be part of UG and thus makes the presence of SVCs plausible.

Other authors have put forward other proposals. Collins (1997) for example proposed that

SVCs obligatorily contains internal argument sharing and there is the existence of an empty

category as an argument of V2 which is coindexed with the object of V1, hence the direct

object of the first verb is understood as the direct object of the second verb. In structures

where there is no direct sharing, at LF the second verb incorporates into the first, i.e. V2

covertly raises and adjoins to V1, thus allowing an account of how embedded verbs in an SVC

can have the subject of the sentence as their external argument.

Aboh (2009), based on facts regarding verb serialization in the Kwa languages, proposed that

the argument-sharing hypothesis cannot be maintained for all SVCs; that it is inaccurate and

must therefore be rejected. He views SVCs as an instance of Object Verb Construction. He

further proposed that SVCs do not involve embedding two or more finite verbs under a single

vP, but alternatively that the two verbs belong to two different domains of the clause

structure: V2 merges as the head of the lexical field of which V1 realizes a functional head

within the extended functional field, hence V1 always precedes V2. Cross-linguistic variation

in SVCs derived from the interaction between object movement and verb movement leads to

V1-XP-V2 versus V1-V2-XP sequences. If this analysis is adopted for the current data, it would

imply that children acquiring JC from an early age are aware of movement rules, and also the


261

distinctions between functional (verbal) elements and their lexical cognates in setting the

serializing parameter.

Regardless of the varying proposals, it is clear that verb serialization is a phenomenon that is

present in early JC.

We will examine the acquisition of Inclusive/Plural marking in the following section.

9.4 INCLUSIVE/PLURAL MARKING

The basilectal plural-marker in JC is post-nominal dem. It is restricted to definite NPs, and

strongly tends to appear in NPs containing the definite article di. Though dem normally

attaches to the pluralized noun directly, it occasionally occurs after an embedded noun

structure as in (60) (Patrick, 2007).

60) A she haffi carry di box a liquor-dem from downstairs?

HL38

3SG MOD carry DEF box of liquor-PL from downstairs

“Is it her job to carry [the boxes [of liquor]] from downstairs?”

(Sistren 1986:231 cited in Patrick 2007)

According to Stewart (2006) dem does not express plurality in the same way that the plural

marking morpheme does in a number-marking language such as English. Dem is not

necessary in JC to express plurality (other ways of expressing plurality will be mentioned

later). When dem is used, it gives group rather than individual reference, and presupposes

38 The pre-posed focus article a is here referred to as HL meaning ‘highlighter’.


262

number only insofar as by definition, a group denotes plurality. Thus in di poliis dem (in (61)

below), dem indicates that it is the totality of police which is under consideration. Its plural

construct is a consequence of it being a marker of inclusiveness (inclusiveness being a feature

of definiteness) – a group of police must necessarily denote (semantic) plurality.

61) Di poliis dem kyari wan gon.

DEF police INCL carry IND gun

“The policemen (together) carried a (i.e. one) gun.” (Stewart 2006)

Also post-nominal dem is also used in constructions such as Mieri dem (Mary INCL) where it

receives the interpretation “Mary and her friends” or, perhaps, “Mary and her family”,

depending on the context. Here dem could not reasonably be said to be marking the plurality

of a proper name, but instead an analysis in terms of inclusiveness, where reference is to a

group of persons in which Mary is included (Stewart, 2006). She therefore proposed that dem

be classified as a marker of inclusiveness, as will be adopted for the current work.

It is well documented that bare nouns in JC may be interpreted as generic singular (62), and

without the inclusive marker dem, as plural (63). These readings are attested in the utterances

of children acquiring JC:

62) Poliis gaan tek mi we. (SHU 3;01)

Police PROS take 1SG away

“Police is going to take me away.”

63) Mi fried a dopi. (TYA 2;11)


263

1SG afraid of ghost

“I am afraid of ghosts.”

In the absence of an inclusive marker, plurality may be also signalled by way of context,

relying on prior discourse information to disambiguate the number of the noun in question:

64) Wi a chaaj di bachri lieta? (RJU 2;06)

1PL PROG charge DET battery later

“Are we going to charge the batteries later?”

Additionally plurality may be signalled by a numeral or a quantifier (65 and 66 respectively)

or by affixation of a –s or –z, as borrowed from English (67):

65) Mi get tuu baik yeside. (COL 2;07)

1SG get two bike yesterday

“I got two bikes yesterday.”

66) Som guot de ya. (KEM 2;09)

Some goat COP LOC

“Some goats are here.”

67) Im Ø get we an iit out di egs. (ALA 2;11)

3SG Ø get away CONJ eat out DET egg-PL

“He will get away and eat out the eggs.”


264

In the present section, we will be examining the development of plurality marking by means

of the overt post-nominal inclusive marker as demonstrated in (68) to (72).

68) A im fren dem. (COL 2;03)

FOC 3SG friend INCL

“They are his friends.”

69) Si i rakstuon dem de. (KEM 2;07)

See DET stone INCL LOC

“There are the stones.”

70) Wa apm tu ar nuoz dem? (RJU 2;07)

What happen to 3SG nose INCL

“What happened to her nostrils?”

71) Som a i stika dem kom aaf. (SHU 2;11)

Some of DET sticker INCL come off

“Some of the stickers came off.”

72) Mi aa wet i biid dem. (TYA 2;08)

1SG PROS wet DET bead INCL

“I am going to wet the beads.”

As shown above, the inclusive marker is an independent morpheme, identical to the 3rd

person

plural pronoun:


265

73) Mi naa_go toch dem. (ALA 3;00)

1SG NEG~PROS touch 3PL

“I am not going to touch them.”

To get the plurality interpretation, it must be post nominal. It co-occurs only with nouns

realized with the definite determiner or with a possessive pronoun.

Table 5 gives the overall distribution of inclusive dem in our corpus. The data demonstrate

that children used the overt post-nominal inclusive marking sparingly in their spontaneous

production. Only 72 utterances were detected where plurality was overtly marked. SHU and

TYA hardly ever used the marker (8 and 3 times respectively) while it is literally unattested in

the productions of ALA.


< 1.59 - 0 0 0 - - 0

1.5 - 2-49 0 0 0 0 2 - 2

2.5 - 3.49 8 0 1 2 2 1 14

3.5 - 4.49 11 0 2 0 2 4 19

4.5 > - 0 11 1 22 3 37

TOTAL 19 0 14 3 28 8 72

Table 5: Overt Post-nominal Inclusive Marker

While it was first attested at 32 months for TYA, it was attested quite early in the production

of COL, RJU, KEM at 27 months of age and at 26 months of age in SHU. This is in

accordance to findings of English, where Brown (1973) reported that children begin

producing plural marker on nouns in 90% obligatory contexts between 24 and 34 months of

age. In an elicitation task, Ferenz and Prasada (2002) found that children consistently marked

the plural appropriately by 27 months. The early acquisition of plural marking appears to be


266

quite universal as reported for other languages: Marrero and Aguirre (2003) for Spanish; Van

Wikj (2007) for Dutch; Szagun (2001) for German; Leonard, Caselli and Devescovi (2002)

for Italian.

Closer analysis of the plural forms found in the corpus reveals that children hardly ever

produced deviant forms during the period under examination. Except for 2 utterances (74 and

75) post-nominal dem never occurs with generic plural nouns, nor was found to co-occur with

numerals or quantifiers without the use of the definite pre-nominal determiner di:

74) *tuu bord dem. (COL 2;03)

two bird INCL

“The two birds.”

75) *A waahn si tuu lait dem. (SHU 3;01)

1SG want see two light INCL

“I want to see the two lights.”

It appears then that children are aware of the rule restricting the use of the overt inclusive

marker in utterances where a plural modifier is already present. This finding is not surprising,

as it is documented in the literature that children acquiring German, produced only 2 to 3%

target-inconsistent plural forms in longitudinal spontaneous data on L1 acquisition (Kopcke,

1998). He proposed that at age three and younger, most of the plural forms are still learned by

rote, and as such a productive pattern of plural formation does not seem to be established as

yet. Unlike for English, irregular noun pluralization is non-existent and as such children do


267

not go through a stage of over-regularization with plural formation. From the beginning,

children acquiring JC demonstrate adult-level knowledge of plural forms.

We will now examine Pronouns and Reflexives in the production of children acquiring JC.

9.5 PRONOUNS AND REFLEXIVES

As evidenced in the data, personal pronouns in JC do not mark a difference between

nominative and accusative case:

76) Im kil im. (COL 2;07)

3SG kill 3SG

“He killed him.”

This is a typical feature of pronouns in creole languages. However a distinction between these

cases and the possessive is optionally marked with the use of fi, as exemplified by the

children’s speech below:

77) I de pan mi fut. (TYA 3;00)

3SG LOC on 1SG foot

“It is on my foot.”

78) Mi wahn kum ova fi yo yaad. (KEM 3;00)

1SG want come over POSS 2nd

sg yard

“I want to visit your home.”


268

79) Fi mi fingga mash wen mi a shot i. (ALA 2;08)

POSS 1SG finger mash when 1SG PROG shut 3SG

“My finger mashed while I was shutting it.”

The reflexive pronouns are also distinct from the personal pronouns as they are followed by

the morpheme self ‘self’. A look at the spontaneous data shows that all 6 children have

reflexives in their grammar, but it was never used in a complete phrase by TYA. Despite its

attestation, the production of the reflexive was sparse in the children’s utterances. Only 66

utterances were produced, as detailed in Table 6 and exemplified in (80) – (83) below.


< 1.59 - 0 0 0 - - 0

1.5 - 2-49 0 2 0 0 0 - 2

2.5 - 3.49 3 5 2 0 0 8 18

3.5 - 4.49 3 2 1 4 1 6 17

4.5 > - 2 7 1 12 7 29

TOTAL 6 11 10 5 13 21 66

Table 6: Production of Reflexive Pronouns

80) Im muuv imself. (COL 2;08)

3SG move 3SG~REF

“He moved himself.”

81) Mi a bai ais-kriim fi maiself. (KEM 2;11)

1SG PROG buy ice-cream for 1SG~REF

“I am buying ice-cream for myself.”

82) A mii a jraiv miself. (RJU 3;00)


269

FOC 1SG PROG drive 1SG~REF

“It is me who is driving myself.”

83) Mi kyaahn manij miself. (SHU 2;08)

1SG can~NEG manage 1SG~REF

“I cannot manage myself.”

The data reveals that in contexts requiring a reflexive pronoun, children would normally

produce such reflexive pronoun in a target-consistent manner. There was only one utterance

where an obligatory reflexive pronoun was omitted yielding ungrammaticality:

84) *Mii wash mi bak fi mi Ø. (TYA 2;08)

1SG wash 1SG back for 1SG~Ø

“I am going to wash my back for myself.”

Additionally, as seen in (85) below, children make no mistakes in binding the reflexive

pronoun with the appropriate antecedent thereby demonstrating adult-like performance and

knowledge of ‘Binding Principle A’: reflexives must be locally bound, i.e. they must have a

local antecedent. Though examples of this sort (with competing antecedents) are sparse in the

dataset, (85) provides marginal evidence that from an early age children are aware of the

principles governing syntactic binding and c-command.

85) Yuu tek aaf i yoself. (ALA 2;06)

2SG take off 3SG 2SG~REF

“You should take it off yourself.”


270

In addition, children also demonstrate knowledge of ‘Principle B’: Pronouns are locally free.

As seen in (86) for the pronoun ar to be locally free it cannot refer to the preceding NP i daag

nor the preceding pronoun mi. Ar can however be bound to shi, as here it is locally free: the

antecedent shi is not in the same clause that contains the pronoun ar thus they may be bound

non-locally. Also we see where the pronoun ar is 3SG and can not be co-indexed with any NP

or other pronoun but the corresponding 3SG root subject. Likewise as demonstrated in (87),

though all three pronouns are 3SG, it cannot be co-indexed with shi as they are in the same

local domain and must be free locally. Shi however can be bound to ii as here ii is clause

external.

86) Shii se mi a_go luus i daag pan ari. (SHU 3;00)

3SG say 1SG PROS loose DET dog on 3SG

“She said that I am going to loose the dog on her.”

87) Iii se shii no waa sii it. (SHU 2;07)

3SG say 3SG NEG want see 3SG

“She said that she does not want to see it.”

The data presented above shows that unlike pronouns, reflexives are always locally bound,

demonstrating children’s knowledge of c-command and Principle A. This is in keeping with

findings for Seselwa and Morisyen-speaking children as presented by Adone (2012). This is

also true for other languages and shows that Creole-speaking children behave in a similar way

when compared to children who speak non-creole languages. On the other hand, there is a

widespread view that young children do not have knowledge of Principle B (see Lust 1986 for

a review), as they tend to bind pronouns locally. Our spontaneous data do not seem to provide


271

evidence of this violation. We however do not have reliably quantitative evidence that

children necessarily always obey Principle B as such complex constructions are scarce in the

data. We therefore are unable to conclude definitively whether children master the pronoun-

reflexive distinction. What we can see however is that children at least demonstrate some

knowledge of the binding principles and mark reflexives in line with the target grammar.

We will now turn to the phenomena surrounding the copula in early JC.

9.6 COPULA

It is well documented in studies on early child language acquisition that children would

generally produce utterances omitting the copula (Radford 1990; Becker 2000; Wexler 2000).

Whereas these utterances would be target-inconsistent in non-creoles, for JC they are

grammatical. According to Bickerton (1981), copula omission with adjectivals and nominals

(in heavily superstrate-influenced creoles such as the Indian Ocean Creoles), and the use of a

distinct verb with locatives and nominals (in less superstrate-influenced creoles such as those

of the Caribbean) are typical features of Creole grammar. However many languages

(including Russian, Japanese, Arabic, American Sign Language, etc.) exhibit these features in

varying contexts and as such it is not restricted to solely creole languages.

We will now examine this phenomenon in the present corpus.

88) Mi taai roun a ous. (ALA 2;05)

1SG toy round at house

“My toy is around by the house.”


272

89) Yuu ina kyar. (RJU 2;04)

2SG into car

“You are in the car.”

90) We iina yu bag? (SHU 2;07)

What into 2SG bag

“What is in your bag?

91) Mi kyar doti. (COL 2;10)

1SG car dirty

“My car is dirty.”

92) Kaal Bringl if yo bad! (KEM 3;02)

call Bringle if 2SG bad

“Call Bringle if you are bad!”

93) I nais. (TYA 2;04)

3SG nice

“It is nice.”

The data reveals that copulas are generally omitted with prepositional (88) – (90) and

adjectival (91) – (93) predicates. Whereas these omissions are grammatical for children

acquiring JC and therefore immediately target-consistent, for some non-creole learners, these

systematic errors result in delayed learning of the respective target languages.


273

The data also shows that children acquiring JC, from an early age, generally produce de and a

target-consistently. Arguably, the overt realization of de and a as demonstrated below could

be instances of the use of the copula. Bailey (1966) classifies this use of de as the “locative

verb ‘be’”, while Durrleman-Tame refers to it as the “locative copula”. Cassidy (1961) (cited

in Durrleman-Tame 2008) also points out that “da39

involves the idea of being in a place, thus

could easily have been converted to uses equivalent to those of English ‘be’”. Likewise

Cassidy (1961) writes that a seems to be clearly verbal, and could possibly be a phonetic

reduction of ‘is’, having the exact force as ‘it is’ or ‘there are’. Durrleman-Tame refers to it as

the “equative a” while Christie (1997) simply terms it as the “copula a”.

94) Krombz de pan yu kluoz. (SHU 2;11)

crumbs LOC on 2SG clothes

“Crumbs are on your clothes.”

95) Shi de uom. (ALA 2;06)

3SG LOC home

“She is at home.”

96) Wan de pan mi fingga. (TYA 3;00)

One LOC on 1SG finger

“One is on my finger.”

97) Yu a di poliis? (SHU 3;01)

2SG EQU DET police

39 Phonetic variation of de.


274

“Are you the police?”

98) Yo mada a iidyat. (KEM 3;01)

2SG mother EQU idiot.

“Your mother is an idiot.”

99) Dat a apl, dat a fish. (RJU 2;09)

DEM EQU apple DEM EQU fish

“That is an apple, that is a fish.”

Note however, that unlike English copula, JC ‘copula-like’ de and a cannot occur clause-

internally with adjectival predicates. De is instead restricted to nominal and prepositional

predicates, while a is restricted to only nominals.

100) *Mi kyar de doti.

1SG car LOC dirty

101) *Mi kyar a doti.

1SG car EQU dirty

De and a behaves differently when it comes to constructions having left-peripheral

movement, such as wh-questions, focalization or topicalization (as discussed in Chapter 8).

When front movement is applied to a predicator involved in a locative relationship to the

subject, the locative marker is left in sentence final position as demonstrated in the focalized

utterance (102c) for the declarative in (102):


275

102) a. Im de skuul.

“He is at school.”

b. *A skuul im.

c. A skuul im de.

On the other hand, when front movement becomes applied to a predicator involved in an

equational relationship to the subject, it is imposible to leave the equational marker a in

exposed sentence final position (Pochard & Devonish, 1986).

103) a. Mi a kyaapinta.

“I am a carpenter.”

b. *A kyaapinta mi a.

c. A kyaapinta mi bii. (Pochard & Devonish 1986)

Pochard & Devonish (1986) argued that the relational and linking role of a before equational

predicators blocks it from occurring without being followed by an item with which to express

the relationship. Note that the same restrictions are applied to utterances involving wh-

interrogation: de must be left in sentence-final position (104) while a cannot occur in this

position (105 a.):

104) We im de?

where 3SG LOC

“Where is he?”

105) a. *Wa yu a?

what 2SG EQU


276

b. Wa yu bi?

what 2SG COP

“What are you?”

As seen in (105b.), a copula-like bi is employed for such equative constructions involving

left-peripheral movement. This copula-like use of bi is also evident in the corpus, as

illustrated in (106) – (109):

106) Wa dis bi? (COL 2;05)

what DEM COP

“What is this?”

107) Uu dem bi? (SHU 2;09)

who 3PL COP

“Who are they?”

108) A no stuul i bi, a baik. (KEM 2;09)

FOC NEG stool 3SG COP FOC bike

“It is not a stool, it is a bike.”

109) A dat mi slippaz bi. (TYA 2;11)

FOC DEM 1SG slippers COP

“That is what my (pair of) slippers is.”


277

As revealed, children acquiring JC produce bi only in specific constructions: mainly in direct

interrogatives (106) – (107), and in focalized utterances (108) – (109). In contrast to copula-

like de and a, bi does not occupy a clause internal position. An arguably similar lexical form

of the copula, realized as ete, appears in Mauritian Creole (MC):

110) Ki Zézi Kris ète? (MC: Baker & Syea 1991)

What Jesus Christ be

“Who is Jesus Christ?”

The introduction of this overt copula in MC is clearly a post-creolization change as MC did

not employ an overt copula, in any circumstances, before the latter part of the 19th

century

(Baker & Syea 1991). According to Henri and Abeillé (2007) it is triggered by predicate

extraction and is restricted to specific contexts. The contexts given (ibid.) however do not

seem to apply to the data in the present corpus. This could be as a result of limitations in the

present corpus, or that the forms are produced later in development. Importantly, Baker &

Syea (ibid) pointed out that Haitian Creole (HC) also employs an overt copula ye which is

found in circumstances very similar to MC. The example below shows the use of the copula

ye in a focalized construction:

111) Se dokte Jak ye (HC: Déprez & Vinet 1997)

FOC doctor Jack ye

“Jack’s a DOCTOR.”


278

Examining the corpus, we note that there are 484 utterances with copula-like bi as detailed in

Table 7. They are produced by all informants with the exception of ALA. They are produced

during the same period where copula-like de, a and zero copula are attested.


< 1.59 - 0 0 0 - - 0

1.5 - 2-49 0 0 0 2 1 - 3

2.5 - 3.49 55 0 5 18 0 55 133

3.5 - 4.49 34 0 48 17 3 113 215

4.5 > - 0 4 14 54 61 133

TOTAL 89 0 57 51 58 229 484

Table 7: Production of Copula bi

Further inspection of copula use in the data reveals borrowings from English. There are 180

utterances with English copula, as exemplified in (112):

112) Wat iz dis? (ALA 2;00)

what COP DEM

“What is this?”

These ‘English-like’ copulas were primarily detected in ALA’s files (74%).

Next we will discuss determiners followed by a brief conclusion.

9.7 DETERMINERS

The set of determiners consists of articles and quantifiers, all of which are prenominal. In the

present section we will be focussing on the following determiners:


279

a. Wan ‘a’ indefinite singular

b. Som ‘some’ indefinite plural

c. (D)i ‘the’ definite40

These are exemplified in (113) through to (115) respectively:

113) Wan big chok a kom. (ALA 2;03)

DET big truck PROG come

“A big truck is coming.”

114) Mi a_go bai som gyas fi it. (KEM 2;08)

1SG PROS buy DET gas for 3SG

“I am going to buy some gas for it.”

115) Mi a chuo it ina i gyaabij. (COL 2;00)

1SG PROG throw 3SG into DET garbage

“I am throwing it in the garbage.”

According to Bickerton’s (1981) specific/nonspecific distinction, creoles use zero or

indefinite articles to mark specific from non-specific NPs. The use of the articles however

does not depend solely on the specificity of the referent, but also on whether it can be

presupposed by the speaker that the listener will know which specific referent the speaker has

in mind. Bickerton (ibid.) states that all non-specific referents usually take a zero article,

whether they are presupposed or not (as in 116). All specific referents take the definite article

40

Note that the definite article is used for both singular and plural referents, but importantly with plurals it utilizes the post-nominal inclusive marker, thereby forming the structure di NP dem.


280

if they are presupposed (as in 117). If the referents are specific but not presupposed, then the

indefinite article is used (as in 118).

116) Im gaan bai sasij. (KEM 2;11)

3SG gone buy sausage

“He went to buy sausage(s).”

117) Shi bai di baik. (ALA 2;11)

3SG buy DET bike

“She bought the bike.”

118) Granmaa bai wan spaida slipaz. (SHU 3;03)

Grandma buy DET spider slipper

“Grandma bought a (pair of) spider slipper.

In order to determine if an utterance is presupposed, the discourse context is highly relevant.

For the most part, children acquiring JC seem to conform to the specific/nonspecific

distinctions as Bickerton’s innate mechanism would predict. We however note some

examples which deviate from the expected pattern:

119) Ø Ø gi mi *wahn tii pliiz? (ALA 2;04)

Ø Ø give 1SG DET tea please

“Can you give me some tea please?”

120) Ø Ø luk fi *wahn badi. (COL 2;11)


281

Ø Ø look for DET body

“I am looking for someone.”

121) *Wahn maal. (COL 2;08)

DET marl

“Some marl.”

122) Mi aa bai *wan gyas. (KEM 2;10)

1SG PROS buy DET gas

“I am going to buy some gas.”

123) *Wahn piipl. (RJU 3;00)

DET people

“Some people.”

124) Mi a_go dig *wahn dort dong a Chav. (KEM 3;00)

1SG PROS dig DET dirt down at Chav

“I am going to dig some dirt down by Chav’s (place).”

125) Si *wan waata oova de-so. (TYA 2;11)

See DET water over LOC

“There is some water over there.”

126) Ø kyan get *som rait? (ALA 2;05)

Ø can get DET write


282

“Can I get a (chance to) write?”

127) Momi put *som jakit on mi. (ALA 2;07)

Mommy put DET jacket on 1SG

“Mommy put a jacket on me.”

128) Si *i worm de! (ALA 2;04)

See DET worm LOC

“There is a worm!”

129) Mi wahn go iina *a bak. (RJU 3;00)

1SG want go into DET back

“I want to go in the back.”

130) Mi wahn *a titi man. (SHU 2;02)

1SG want DET breast COM

“I want breast.”

131) A ier *i myuuzik. (COL 2;08)

1SG hear DET music

“I heard music.”

In (119) to (125), we see where the children substitute the indefinite singular wan for the

indefinite plural form som and vice versa in (126) and (127). In (128) the child produced the

definite article i instead of the indefinite wan and vice versa in (129). (130) and (131) show


283

the use of the indefinite and definite articles respectively in utterances where zero article

should be employed. One could argue that some of these utterances are well-formed in the

target language; however their target inconsistency is primarily based on the context of

utterance and the intended meaning. Being that there are only 13 such utterances in the entire

corpus indicates that these types of errors are infrequent and do not appear to be a true

characteristic of the grammars of the children. Marastos (1976) attributes these errors to the

fact that children occasionally fail to keep track of previous unspecified referents rather than

them being a display of general ignorance of the definite-indefinite and specific-non-specific

distinctions.

One area that does not fit nicely in the classification of articles based on Bickerton’s

specificity and presupposedness rules is the naming or identification of nouns. According to

Brown (1973), named referents are specific but not presupposed and as such should take the

indefinite article. On the other hand Maratos (1976) argues that named referents may

sometimes be specific and presupposed, and therefore are problematic for the classification of

articles. For creoles, Bickerton claims that they use zero articles for naming, however

Valdman (cited in Adone 1994) reports that an indefinite article is used in such contexts.

Adone (1994) reports that most of her Mauritian Creole informants prefer to use a zero article,

hence supporting Bickerton, however there are also speakers who choose to use the indefinite

article. The data from the present study shows that JC children would globally prefer the use

of a zero article as in (132):

132) INV: A wa dat?

“What is that?”

CHI: Baik. (COL 1;09)


284

bike

“(It is a) bike.”

INV: Yaa_go luk, we yo si out de?

“You’re going to look, what do you see out there?”

CHI: Van. (COL 1;09)

van

“(I see a) van.”

However, though minimally attested, the use of the indefinite article is not impossible:

133) INV: A we it bi?

“What is it?”

CHI: Wahn man. (COL 2;09)

DET man

“(It is) a man.”

The general data shows that children acquiring determiners in JC are for the most part target-

consistent from the beginning. In addition the use of the zero article for nominal identification

is in line with Bickerton’s predictions.

9.8 CONCLUSIONS

The primary purpose of this chapter was to describe the acquisition of some syntactic

constructions which are generally characterized as being typical creole features and to

evaluate their development in terms of errorless acquisition. We examined the production of


285

negation, verb serialization, inclusive/plural marking, pronouns and reflexives, copula and

determiners.

Cumulatively, the data provide evidence that, except for a few performance errors, children

acquiring JC generally use the appropriate forms from the start, suggesting immediate target

consistency. The early age in which the constructions are attested is strong indication that

these features are possibly part of UG. Additionally, the fact that similar constructions are

reported in the production of children acquiring non-creole languages (though some of which

are target-inconsistent), demonstrates a systematic universal pattern of development. Being

that the structures studied in this section are not atypical to many non-creole languages, the

findings are supportive of a theory of universality in the early stages of language acquisition,

rather than that creoles are exceptional languages. Some of the structures examined in this

chapter were limited in the spontaneous productions of the children, thus pointing towards the

need for additional studies employing experimental methodologies.

Chapter 10: Conclusions and Implications

286

CHAPTER 10

CONCLUSIONS AND IMPLICATIONS

10.1 INTRODUCTION

The research discussed in this dissertation serves to fill a gap in acquisition studies with

respect to creole languages by elucidating the acquisition of JC syntactic structures. The

mapping of these structures onto the target system has revealed that the production of children

acquiring JC is mainly adult-like with the exception of target-inconsistent omissions. These

omissions are however highly systematic and consistent with UG. Additionally, the general

discussions argue in favor of a full competency approach to development. The data analyzed

herein will be made accessible on the CHILDES Database and thus facilitate future work on

Creole acquisition.

10.2 SUMMARY AND CONCLUSIONS

The data presented in this dissertation points to a system in which children possessed both

lexical and functional structures co-occurring at all levels of acquisition. This is in contrast to

Radford’s (1990) proposal that functional structures are globally delayed with respect to

lexical structures. As detailed in Chapter 5, we observed the gradual expansion of both lexical

and functional words/morphemes. The general quantitative expansion of the various items

within the grammatical categories points to a system of syntactic acquisition that is gradual,

rather than reflecting immediate categorical introduction and/or eradication of the previous

grammatical system. The steady increase in words/morphemes in contrast to lexical and

functional categories shows that all the grammatical heads are present from the onset but


287

mature overtime. So while there were on-going introduction of novel lexemes within the

respective domains, the grammatical categories themselves (i.e. nouns, verbs, adjectives,

determiners, inflections, complementizers, etc.) remained constant.

Findings from the discussions show that from the very first combinations, children acquiring

JC fix parametric values governing distribution correctly. So the order in which paired

constituents (e.g. heads and their complements) can occur was fixed from within the two-

word combinations stage. Also there were no errors with regards to the distribution of

functional items such as determiners, negation, etc.; thus the distribution of word order and

functional morphology was always target-consistent. Moreover, from an early age, children

displayed knowledge of the rich cartographic hierarchy as attested in the target language and

do not entertain deviant structures. Recall that it was shown that children do not make

mistakes in the combination of TMA markers, nor were there any errors in the relative

ordering of various elements in the CP. The incremental structure building approach to

development therefore cannot account for the early co-occurrence of elements located at

different levels in the clausal structure.

The data further revealed that children acquiring JC produce certain structures without errors

thereby reflecting immediate target consistency, whereas children learning non-creole

languages have been reported to produce similar constructions yielding target inconsistency.

The fact that children, regardless of their target language, produce similar utterances

demonstrates a systematic universal pattern of development. This points to default values for

parameters of UG which are later modified on the basis of input. While this finding seems to

be in line with Bickerton’s ideas, it does not argue in favor of a system that upholds creoles as

being exceptional languages. Moreover, features formerly attested to be typical of creole


288

languages are present in many non-creoles and as such cannot be held as identifying

characteristics of creole languages.

Additionally, we see where children acquiring JC, like children acquiring non-creole

languages, target-inconsistently omit certain elements. As evidenced in the robust production

of null subjects, the omission of wh-words in constituent questions, the omission of focus

markers or the dropping of obligatory aspectual markers, these omissions are consistently and

systematically governed. Children are therefore knowledgeable of the optionality of an overt

spell-out of clausal boundaries made available to them by UG. When this option ceases to be

operative around the end of the third year of life, omissions decline and the child grammar is

basically adult-like. These systematic omissions detected in the speech of children acquiring

JC are universal and thus do not lend support to the creole exceptionality hypothesis.

The cumulative facts from JC as investigated in this dissertation therefore argue in favor of a

full competence approach to linguistic acquisition and development. Overall the findings are

indicative of language universals and argue in support of UG.

10.3 IMPLICATIONS FOR FUTURE WORK

This research led to some important findings that have implications for future research. Firstly

the corpus is limited in providing morpho-syntactic coding of only the children’s speech and

as such many well-needed analyses of the input data are not investigated. It would be

profitable in making claims about language use and developmental stages in acquisition to

establish a comparative analysis of the phenomenon with reference to adult language use, in

order to establish what is specific to child language. It is therefore imperative for further


289

research to be conducted, looking particularly at the contributions of the input language in

creole acquisition.

Notwithstanding the success of the longitudinal approach to the data collection, many

grammatical structures are not sufficiently produced in contexts of spontaneous speech and as

such do not provide an effective quantitative basis for analysis. In order to adequately and

reliably map the entire development of JC, undertaking additional research from an

experimental perspective would be profitable.

Furthermore, it is important in language acquisition studies to examine not only the specific

grammatical production, but also to examine the semantic functions that these categories fulfil

in the child’s system. In order to understand why certain features are more or less frequent

across certain stages of development, research should be structured so as to capture the

communicative need and context of use. Such an approach can only be successful if studies

on language acquisition contain a component whereby the context of utterance is coded and

duly included in the analyses.

Finally, this dissertation has contributed to filling gaps that exist in acquisition research and

enables a better understanding of some of the theoretical issues with which acquisitionists

have been grappling for decades. Whether the conclusions may be generalized to other creole

languages should be further investigated. It is hoped that this work will lead to comparable

research into the acquisition of syntactic structures of other creoles and will help us to

increase our knowledge of how the general mechanisms governing languages are acquired.

… let this not be the end, but the beginning of greater things to come.

290

Appendix

Appendix 1: Consent Form

Consent Form for Participation in Research

Study Title: Acquisition of Jamaican Creole Syntax: A corpus-based study of early parameter

setting

Principal Investigator: Tamirand De Lisser

Department of Linguistics

University of Geneva

Purpose of this Study: The purpose of the study is to gather data in order to explore the

emergence and early acquisition of the syntax of Jamaican Creole.

Procedures: Research will be carried out in three phases: Data collection, analysis of the data

and theoretical discussion. Your child/ward will be involved in the initial phase; which will

entail the construction of a corpus of natural production by 6 children: Each child will be

recorded for 45-60 minutes, 3 times per month for a period of 18 months. The recorded

material will be immediately transcribed, in accordance with the CHAT system guidelines of

the CHILDES (Child Language Data Exchange System) Database.

291

In the second phase of the study, this corpus will be subjected to detailed analysis, aimed at

describing both target-consistent and target-inconsistent production.

The third phase of this project will integrate the empirical findings obtained in the corpus

study (phase II) into the theoretical discussion of early parameter setting and evaluate

competing accounts of early syntactic development.

Participant Requirements: Children, approximately 18 months old, from households where

the syntactic features of Jamaican Creole are predominant are eligible to participate in this

study.

Risks: There are no risks associated with participation in this study which is non-invasive.

Benefits: The recording sessions will be conducted in a playful manner, such that the child

will feel inclined to interact verbally. As a result, the sessions should be enjoyable for the

participating child and allow for some free time for you, the parents/guardians. Moreover, by

participating you will allow us to improve our understanding of Creole language acquisition

and we will be able to give you information regarding your child’s strengths and weaknesses

in language acquisition.

Confidentiality: By participating in the study, you understand and agree that we may be

required to disclose your consent form, data and other personally identifiable information as

required by law, regulation, or court order. Otherwise, confidentiality will be maintained in the

following manner:

Your child’s/ward’s data and consent form will be kept separate. Your consent form will be

stored at the University of Geneva and will not be disclosed to third parties. By participating,

292

you understand and agree that the data and information gathered during this study may be used

and published; however, name, address, contact information and other direct personal identifiers

in your consent form will not be mentioned in any publication or dissemination of the research

data and/or results.

The researchers will take the following steps to protect participants’ identities during this

study: (1) All research data will be assigned a participant code; (2) The researchers will

record any data collected during the study by participant code, not by name; (3) Original

recordings and data files will be stored in a secured location accessed only by authorized

researchers; (4) Access to the audio and/or video language transcript database is restricted to

only authorized researchers.

Optional Permissions: I understand that the researchers may want to use a short portion of

any video or audio recording for illustrative reasons in presentations of this work for scientific

or educational purposes. I give my permission to do so provided that the data will be

anonymous.

YES NO (Please initial here ____)

Rights: Your participation is voluntary. You are free to stop participation at any point. Refusal

to participate or withdrawal of your consent or discontinued participation in the study will not

result in any penalty.

The researcher may at his/her discretion remove you from the study for any reason. In such

an event, you will not suffer any penalty.

293

Voluntary Consent: By signing below, you warrant that you have the authority to make

decisions on behalf of the participant. You agree that the above information has been

explained to you and all your current questions have been answered. You understand that

you may ask questions about any aspect of this research study during the course of the

study and in the future. By signing below, you consent to the participant’s involvement in

this study.

___________________________________

NAME OF PARTICIPANT

___________________________________ ____________________________

SIGNATURE OF PARENT/GUARDIAN DATE

I certify that I have explained the nature and purpose of this research study to the above

individual and I have discussed the potential benefits and possible risks of participation in

the study. Any questions the individual has about this study have been answered and any

future questions will be answered as they arise.

___________________________________ ____________________________

SIGNATURE OF PERSON OBTAINING CONSENT DATE

294

Appendix 2: Codes and grammatical relations of Jamaican Creole

CODES GRAMMATICAL RELATIONS EXAMPLES

PREDICATE RELATIONS

T:ant Anterior tense did/ben/ en

T:fut Future tense wi

Mod:epis Epistemic or possibility modal shuda, maita

Mod:nec Necessity modal mos

Mod:abl Ability or permission modal kyan

Mod:obl Obligation modal afi

Mod:pred Prediction modal wi

Mod:vol Volition modal wi

Asp:prog Continuative/ progressive aspect A/de/da

Asp:comp Completive or terminative aspect don

Asp: retro Retrospective aspect jos

Asp:pros Prospective aspect a go/ gaa

V:sta Stative verb nuo

V:nonst Non-stative verb gi

INF Infinitive fi

NEG Negator no

V:SVC:ins Instrumental serial verb Tek naïf kot bred

V:SVC:dir Directional serial verb Ron kom gi mi

V:SVC:dat Dative serial verb Bai at gi mi

V:SVC:ben Benefactive serial verb Pap stuori gi mi

V:SVC:cau Causative serial verb Mek glaas brok

295

V:SVC:int Intentional (or purpose) serial verb Go taak se mi lai

V:SVC:res Resultative serial verb Shat bod ded

V:loc Locative verb de

PRED:Adj Predicative adjective priti

SUBJECT RELATIONS

Pro:1s 1st person singular pronoun mi

Pro:2s 2nd person singular pronoun yu

Pro:3s 3rd person singular pronoun im

Pro:1p 1st person plural pronoun wi

Pro:2p 2nd person plural pronoun unu

Pro:3p 3rd person plural pronoun dem

Pro:wh Wh- pronoun we

Pro:dem Demonstrative pronoun dat

REF Reflexive pronoun -self

N:prop Proper noun Jan

N:com Common noun buk

N:abs Abstract noun chuut

POSS Possessive fi

Det:def:s Definite singular determiner di

Det:def:pl Definite plural determiner dem

Det:dem:s Singular demonstrative adjective Dis/dat

Det:dem:p Plural Demonstrative adjective dem

Q:indef:s Indefinite singular quantifier wan

Q:indef:pl Indefinite plural quantifier som

296

Q:NUM Numeral Faiv

Incl Inclusive/Plural marker dem

ADV:loc Locative adverb de

ADV:temp Temporal adverb yeside

ADV:asp Aspectual adverb aredi

ADV:deg Degree modifier veri

Foc focus a

Equ Equative copula a

ADJ Adjective lang

Cop Copula bi

EXTRA-CLAUSAL ELEMENTS

CONJ:cond Conditional conjunction den

CONJ:coo Coordinating conjunction ahn

CONJ:subor Subordinating conjunction bot

Prep Preposition pan

COMP Complementizer dat

REL Relativizer we

COM Communicator eloo

UNK Unknown xxx

ONO Onomatopoeia Bow-wow

DATE Date May 25, 2013

STRUCTURES FROM ENGLISH

Asp:prog-eng English Progressive Aspect -ing

297

Incl-eng English Plural -s/z

Cop-eng English Copula iz

Poss-eng English Possessives main

OTHERS

0subj Null obligatory subject

0obj Null obligatory object

0wh Null obligatory wh

0V Null obligatory verb

0 Null elements

. Declarative

! Imperative

? Interrogative

+ANT Unmarked Past reading

-ANT Unmarked Present reading

298

APPENDIX 3: Distribution of TMA in the Input

INPUT TO PHASE 1


AGE RANGE 1;8,17 - 1;11,28

1;9,25 - 2;1,20

1;10,4- 2;0,30

1;11,25 - 2;7,23

2;0,21 - 2;4,16 -

1;8,17 - 2;7,23

ASP COMP 3 8 17 12 10 - 50

ASP PROS 200 276 207 597 272 - 1552

ASP PROG 964 970 850 1855 1048 - 5687

ASP RETRO 9 19 18 31 14 - 91

MOD ABL 147 202 155 321 140 - 965

MOD OBL 15 28 22 33 34 - 132

MOD NEC 26 30 4 33 18 - 111

MOD EPIS 18 5 5 24 14 - 66

MOD WI 6 15 16 17 45 - 99

PAST TENSE 34 63 22 91 35 - 245

PAST PROG 10 30 16 23 22 - 101

TOTAL 1432 1646 1332 3037 1652 0 9099

INPUT TO PHASE 2


AGE RANGE 2;0,12 - 2;6,25

2;2,6 - 2;5,7

2;1,15 - 2;6,18

2;8,8 - 2;10,16

2;5,2 - 2;6,20

2;1,23 - 2;6,20

2;0,12 - 2;10,16

ASP COMP 10 5 13 8 5 17 58

ASP PROS 351 215 326 211 121 318 1542

ASP PROG 1687 755 1336 655 466 1358 6257

ASP RETRO 20 27 20 6 6 23 102

MOD ABL 239 229 335 106 66 236 1211

MOD OBL 44 41 37 17 10 55 204

MOD NEC 22 22 14 5 12 37 112

MOD EPIS 22 26 28 9 6 15 106

MOD WI 46 28 88 10 13 23 208

PAST TENSE 109 81 106 37 12 80 425

PAST PROG 27 34 43 13 20 24 161

TOTAL 2577 1463 2346 1077 737 2186 10386

299

INPUT TO PHASE 3


AGE RANGE 2;7,8 - 2;11,7

2;5,23 - 3;0,15

2;7,5 - 3;0,25

2;11,0 - 3;2,15

2;7,5 - 3;3,11

2;7,16 - 3;4,13

2;5,23-3;4,13

ASP COMP 9 24 24 14 29 36 136

ASP PROS 225 460 385 281 544 578 2473

ASP PROG 1084 1617 1578 873 2097 2469 9718

ASP RETRO 9 53 16 14 23 47 162

MOD ABL 217 421 330 197 487 475 2127

MOD OBL 40 89 38 27 62 122 378

MOD NEC 10 42 13 15 41 59 180

MOD EPIS 16 43 23 20 27 39 168

MOD WI 29 71 87 6 110 69 372

PAST TENSE 58 156 99 23 132 113 581

PAST PROG 18 58 83 7 42 44 252

TOTAL 1715 3034 2676 1477 3594 4051 16547

300

REFERENCES

Aboh, E. O. 2009. Clause Structure and Verb Series. Linguistic Inquiry, Vol. 40. No. 1. 1 –

33.

Aboh, E. O. 2004. The Morphosyntax of Complement-Head Sequences: Clause Structure and

Word Order Patterns in Kwa. New York: Oxford University Press.

Adone, D. 1994. The Acquisition of Mauritian Creole. Amsterdam: John Benjamins.

Adone, D. & A. Vainikka. 1999. Long distance Wh-movement in Child Mauritian Creole, in

M. Degraff (ed.), Creolization, Language Change and Language Acquisition.

Cambridge, MA: MIT Press, 75-94.

Adone, D. 2012. The Acquisition of Creole Languages: How Children Surpass their Input.

NewYork: Cambridge University Press.

Aissen, J. L. 1992. Topic and Focus in Mayan. Language, Vol. 68, No.1. 43 – 80.

Allen, S. E. M. 2000. A discourse-pragmatic explanation for argument representation in child

Inuktitut. Linguistics, 38 (3), 483 – 521.

Andersen, R. W. & Y. Shirai. 1996. The primacy of aspect in first and second language

acquisition: The pidgin-creole connection. In William C. Ritchie & Tej K. Bhatia

(eds.), Handbook of second language acquisition. New York: Academic Press. 527- 571.

Antinucci, F. & R. Miller. 1976. How children talk about what happened. Journal of Child

Language 3. 169 - 189.

Bailey, B. 1966. Jamaican Creole Syntax: A Transformational Approach, Cambridge: CUP

Baker & Syea. 1991. Copula in Mauritian Creole. In F. Bryne & T. Huebner (eds.),

Development and structures of Creole Languages: Essays in Honor of Derek Bickerton.

Amsterdam: John Benjamins.

Bar-Shalom, E. 2002. Tense and Aspect in Early Child Russian. Language Acquisition 10,

No.4. 321 - 337.

301

Bates, E., V. Marchman, D. Thal, L. Fenson, P. Dale, S. Reznick, J. Reilly & J. Hartung.

1994. Developmental and stylistic variation in the composition of early vocabulary.

Journal of Child Language, 21(1), 85-124.

Bayer, J. 2010. Wh-drop and recoverability. In Zwart, Jan-Wouter and Mark de Vries (eds.),

Structure Preserved: Studies in Syntax for Jan Koster. Amsterdam: Benjamins.

Becker, M. 2000. The Development of the Copula in Child English: The Lightness of Be.

PhD Dissertation. University of Pennsylvania.

Bellugi, U. 1967. The acquisition of the system of negation in children’s speech. Ph.D.

Dissertation, Harvard University.

Bellugi, U. 1971. Simplification in children’s language. In R. Huxley & E. Ingram (eds.),

Language Acquisition: Models and Methods. New York: Academic Press.

Berwick, R. 1982. Locality Principles and the Acquisition of Syntactic Knowledge, PhD

dissertation, Department of Computer Science and Electrical Engineering, MIT,

Cambridge, MA.

Bianchi, V., G. Bocci and S. Cruschina. 2013. Focus fronting and its implicatures. In E. Aboh

et al (eds.) Romance Languages and Linguistic Theory 2013: Selected Papers from

Going Romance, Amsterdam 2013. Amsterdam: John Benjamins.

Bickerton, D. 1984. The language bioprogram hypothesis, Behavioural and Brain Sciences 7:

173–188.

Bickerton, D. 1981. Roots of Language. Karoma Publishers, Ann Arbor.

Bickerton, D. 1999a. How to acquire language without positive evidence: What

Acquisitionists can learn from Creoles. In M. Degraff (ed.), Language Creation and

Language Change. Cambridge, MA: MIT Press. 49 - 74.

302

Bickerton, D. 1999. Creole Languages, The Language Bioprogram Hypothesis, and Language

Acquisition. In Ritchie & Bhatia (eds.), Handbook of Child Language Acquisition. New

York: Academic Press. 195 – 220.

Bloom L. 1970. Language development: Form and function in emerging grammars.

Cambridge, MA: MIT Press.

Bloom, L., P. Miller & L. Hood. 1975. Variation and reduction as aspects of competence in

language development. In A. Pick (ed.), The 1974 Minnesota Symposia on Child

Psychology, 9. 3 – 55. Minneapolis: University of Minneapolis Press.

Bloom, L., Lifter, K., & J. Hafitz. 1980. The semantics of verbs and the development of verb

inflections in child language. Language, 56, 386-412.

Bloom, L., S. Merkin & J. Wotten. 1982. Wh-Questions: Linguistic factors that contribute to

the sequence of acquisition. Child Development, Vol. 53, No. 4. 1084 – 1092. Wiley.

Bloom P. 1990. Subjectless sentences in child language. Linguistic Inquiry 21. 491 - 504.

Bocci, G. 2007. Criterial positions and left periphery in Italian: Evidence for the syntactic

encoding of contrastive focus. Nanzan Linguistics: Special Issue 3, Vol.1. 35-70.

Boland, J. 2006. Aspect, tense and modality: Theory, typology, acquisition. Volume II.

Doctoral Dissertation. LOT. The Netherlands.

Borer, H. & K. Wexler. 1987. The Maturation of Syntax. In T. Roeper amd E. Williams, eds.,

Parameter Setting, Dordrecht: Reidel.

Bowerman, M. 1982. Evaluating competing linguistic models with language acquisition data:

Implications of developmental errors with causative verbs. Quaderni di Semantica 3,

5–66.

Bromberg H. S. & K. Wexler. 1995. Null subjects in Wh-questions. In C. T. Schütze, J. B.

Ganger & K. Brohier (eds.), Papers on Language Processing and Acquisition, MIT

Working Papers in Linguistics 26, 221-248.

303

Bronckart, J. P. & H. Sinclair. 1973. Time, Tense and Aspect. Cognition 2. 107 - 130.

Brown, R. 1973. A First Langauge. The Early Stages. Harvard University Press, Cambridge,

MA.

Brown, R. & U. Bellugi. 1964. Three processes in the child’s acquisition of syntax. Harvard

educational review, 34, 133-151.

Bryan, B. 2004. Language and Literacy in a Creole-dominant environment: A study of

primary schools in Jamaica. Language, Culture and Curriculum 17 (2), 87-96.

Carpenter, K. 2009. Ow wi laan fi taak a yaad: Jamaican parent-child interaction in home

language development. Final report prepared for Caribbean Child Support Initiative.

Cassidy, F. G. 1961. Jamaica Talk: Three Hundred Years of the English Language in

Jamaica. London: Macnillan.

Chiat, S. 1986. Personal Pronouns. In P. Fletcher & M. Garman (eds.), Language Acquisition:

Studies in First Language Development. (2nd edition). Cambridge: C.U.P.

Chomsky, N. 1959. Review of Verbal Behaviour by B.F. Skinner. Language 35:1. 26 – 59.

Chomsky, N. 1981. Lectures in Government and Binding. Dordrecht: Foris.

Chomsky, N. 1995. The Minimalist Program. Cambridge, MA: MIT Press.

Chomsky, N. 2001. Derivation by phase. In M. Kenstowicz (ed.), Ken Hale: a life in

language. Cambridge, MA: MIT Press.

Chomsky, N. & H. Lasnik. 1993. The theory of principles and parameters. In J. von Stechow,

A. Jacobs, W. Sternefeld, & T. Vennemann, (eds.), Syntax: an international handbook of

contemporary research. Berlin: De Gruyter. Reprinted in: The minimalist program, ch.1.

Chomsky 1995, 13-127. Cambridge MA: MIT Press.

Christie, P. 1997. Thematization in Jamaican Speech. Uwiling 3.

Cinque, G. 1999. Adverbs and Functional heads: A cross-linguistic perspective. Oxford

University Press, Oxford.

304

Clahsen, H. 1988. Parameterized grammatical theory and language acquisition: a study of the

acquisition of verb placement and inflection by children and adults. In S. Flynn & W.

O’Neil (eds.), Linguistic theory in second language acquisition. Dordrecht: Kluwer.

Clahsen, H. (ed.) 1996. Generative Perspectives on Language Acquisition. Amsterdam &

Philadelphia: John Benjamins.

Clahsen, H., M. Penke, T. Parodi. 1993/94. Functional Catories in Early Child German,

Language Acquisition 3, 395-429.

Clahsen H., C. Kursawe & M. Penke. 1996. Introducing CP: Wh-questions and subordinate

clauses in German child language. In C. Koster & F. Wijnen (eds.), Proceedings of the

1995 Groningen Assembly on Language Acquisition, GALA. Groningen: The Netherlands.

Clancy, M. Patricia. 1989. Form and function in the acquisition of Korean wh-questions.

Journal of Child Language, 16. 323 - 347.

Collins, C. 1997. Argument Sharing in Serial Verb Constructions. Linguistic Inquiry 28. 461

– 497.

Copley, B. 2002. The semantics of the future. Ph.D. thesis, Massachusetts Institute of

Technology.

Costa J. & F. Pratas. 2012. Embedded null subjects in Capeverdean. Journal of Linguistics. 1-

21. Cambridge University Press.

Costas, J. & K. Szendroi. 2006. Acquisition of Focus Marking in European Portuguese.

Evidence for a unified approach to focus. In V. Torrens & L. Escobar (eds.), The

acquisition of syntax in romance languages. Amsterdam: John Benjamins.

Crain, S. 1991. Language acquisition in the absence of experience. Behavioral and Brain

Sciences 14, 597–650.

Crain, S. & R. Thorton. 2006. The Acquisition of Syntax. In Encyclopedia of Cognitive

Science. Malden Ma: Wiley.

305

Crisma P. 1992. On the acquisition of Wh-questions in French. Geneva Generative Papers

0.1-2: 115-122.

DeCamp, D. 1971. Towards a generative analysis of post-creole continuum. In D. Hymes

(ed.), Pidginization and Creolization of Languages, Cambridge: CUP. 349 - 370.

Degraff, M. 1999. Language Creation and Language Change: Creolization, diachrony, and

development. Cambridge, MA: MIT Press.

Degraff, M. 2003. Against Creole Exceptionalism. Language 79.2.

Degraff, M. 2004. Against Creole Exceptionalism (redux). Language 80.4.

Deprez, V. & A. Pierce. 1993. Negation and Functional Projections in Early Grammar.

Linguistic Inquiry 24, No. 1. 25 – 67.

Déprez, V. & M-T. Vinet. 1997. Predicative constructions and functional categories in Haitian

Creole. Journal of Pidgin and Creole Languages 12. 203 - 136.

Durrleman, S. (2000). The Architecture of the clause in Jamaican Creole, Generative

Grammar in Geneva 1, 189- 241.

Durrleman, S. and U. Shlonsky. 2015. Focus and wh in Jamaican Creole: Movement and

Exhaustiveness. In U. Shlonsky (ed.), Beyond functional sequence. New York: Oxford

University Press.

Durrleman-Tame, S. 2008. The Syntax of Jamaican Creole: A cartographic perspective.

Ansterdam: John Benjamins.

Erbaugh, M. 1992. The acquisition of Mandarin. In D. I. Slobin (ed.), The crosslinguistic

study of language acquisition, Vol. 3. 373–455. Hillsdale, NJ: Erlbaum.

Felix, S. 1980. Cognition and language development: A German child’s acquisition of

question words. Nehls, D. (ed.), Studies in language acquisition. Gross, Heidelberg. 91 -

109.

306

Ferenz, K. S., & S. Prasada. 2002. Singular or plural? Children’s knowledge of the factors

that determine the appropriate form of the count nouns. Journal of Child Language, 29,

49–70.

Fletcher, P. 1985. A child’s learning of English. Oxford: Basil Blackwell/Andre Deutsch.

Francis, C. 2012. Greater Treasure Beach Sustainable Development Plan: 2030 & Beyond

Community Profiles. St. Elizabeth Parish Development Committee, St. Elizabeth Parish

Council, Canadian Urban Institute & National Housing Trust.

Fraser, C., U. Bellugi, & R. Brown. 1963. Control of Grammar in imitation comprehension,

and production. Journal of Verbal Learning and Verbal Behavior. 2, 121 – 135.

Gentner, D. 1982. Why nouns are learned before verbs: linguistic relativity versus natural

partitioning. In S.A. Kuczaj II (ed.), Language Development vol.2 Language, thought,

and culture. Hillsdale, NJ: Lawrence Erlbaum. 301 – 334.

Gilkerson, J., N. Hyams, & S. Curtiss. 2003. On the scope of negation: More evidence for

early parameter setting. Proceedings of GALA 10, LOT, University of Utrecht, The

Netherlands.

Gillette, J., H. Gleitman, L. Gleitman & A. Lederer. 1999. Human simulations of vocabulary

learning. Cognition, 73, 35 – 176.

Gleitman, L. R. & J. Gillette. 1999. The Role of Syntax in Verb Learning. In William C.

Ritchie and Tej K. Bhatia (eds.), Handbook of Child Language Acquisition. Academic

Press, San Diego California. 279 - 295.

Gooden, S. 2014. Aspects of the intonational phonology of Jamaican Creole. In S. Jun (ed.)

Prosodic Typology II: The Phonology of Intonation and Phrasing. United Kingdom:

Oxford University Press.

Greenfield P. & J. H. Smith. 1976. The structure of communication in early language

development. New York: Academic Press.

307

Gualmini, A., S. Maciukaite & S. Crain. 2002. Children’s insensitivity to contrastive stress in

sentences with ‘only’. Proceedings of the 26th Annual PLC. University of

Pennsylvania.

Guasti, M. T. 2000. An Excursion into Interrogatives in Early English and Italian. In M-A.

Friedemann & L. Rizzi (eds.), The Acquisition of Syntax. London: Longman. 105 - 128.

Guasti, M. T. 2002. Language Acquisition: The Growth of Grammar. The MIT Press,

Cambridge, London.

Guasti, M. T. & L. Rizzi. 1996. Null AUX and the acquisition of residual V2. BUCLD 20

Proceedings. 284 – 295.

Guilfoyle E. 1984. The acquisition of tense and the emergence of lexical subjects in child

grammars. McGill Working Papers in Linguistics. Linguistics Department, McGill

University, Montréal, Québec.

Gundel, J. K. & F. Thorstein. 2004. Topic and Focus. In L. Horn and G. Ward (eds.), The

Handbook of Pragmatic Theory. Blackwell. 174 - 196.

Haegeman L. 1995. Root infinitives, tense, and truncated structures in Dutch. Language

Acquisition 4(3). 205 - 255.

Haegeman L. 1996a. Root infinitives and initial root null subjects in early Dutch. In C. Koster

& F. Wijnen (eds.), Proceedings of the Groningen Assembly on Language Acquisition,

239 - 250. Groningen: Center for Language and Cognition.

Haegeman, L. 2000. Adult Null Subjects in non-Pro Drop Languages. In M.A. Friedemann &

L. Rizzi (eds.), The Acquisition of Syntax. Longman, New York. 129 - 169.

Hamann, C. 1996. Null-Arguments in German Child Language. Language Acquisition 5:3.

155-208.

308

Hamann C. 2000a. The acquisition of constituent questions and the requirements of

interpretation. In M-A. Friedemann & L. Rizzi (eds.), The acquisition of syntax: Studies

in comparative developmental linguistics, 170-201. London: Longman.

Hamann C., L. Rizzi & U. Frauenfelder. 1996. On the acquisition of subject and object

pronouns in French. In H. Clahsen (ed.), Generative perspectives on language

acquisition, 310 - 334. Amsterdam: John Benjamins.

Hamann, C. & K. Plunkett. 1998. Subjectles sentences in child Danish. Cognition 69, 35-72.

Haspelmath et al. 2013. Expression of Nominal Subjects. In Michaelis et al (eds.), The Atlas

and Survey of Pidgin and Creole Languages. 244 – 247. United Kingdom: Oxford

University Press

Henri, F. & A. Abeillé. 2007. The syntax of copular constructions in Mauritian. On-line

Proceedings of the HPSG-2007 Conference, Stanford: CSLI Publications.

Hoekstra, T. & P. Jordens. 1994. From adjunct to head. In T. Hoekstra & B. Schwartz (eds.),

Langauge Acquisition Studies in Generative Grammar. Amsterdam: John Benjamins.

Hohenberger, A. 2002. Functional categories in language acquisition. Self-organization of a

dynamical system. Tübingen: Narr.

Huddleston, R. 1995. The case against Future Tense in English. Studies in Language 9: 399

– 446.

Hyams N. 1986. Language acquisition and the theory of parameters. Dordrecht: Reidel.

Hyams, N. 1992. A Reanalysis of Null Subjects. In J. Weissenborn, H. Goodluck & T.

Roper (eds.), Theoretical Issues in Language Acquisition. New Jersey: Lawrence

Erlbaum Associates.

Hyams N. and K. Wexler. 1993. On the grammatical basis of null subjects in child language.

Linguistic Inquiry 24:421-459.

309

Jaeggli, O. and N. Hyams. 1988. Morphological uniformity and the setting of the null subject

parameter. Proceedings of the NELS 18, 239-253.

Jaeggli O. & K. Safir.1989. The null subject parameter and parametric theory. In O. Jaeggli &

K. Safir (eds.), The null subject parameter, 1 - 44. Dordrecht: Kluwer.

Jakubowicz, C and N. Strik. 2008. Scope-marking strategies in the acquisition of long

distance wn-questions in French and Dutch. Language and Speech 51 (1 & 2):101 – 132.

Jumieka Langwij. http://www.jumieka.com/aatagrafi.html. (Accessed September 24, 2014).

Kampen J. Van. 1997. First steps in Wh-movement. PhD Dissertation, Utrecht University.

Kim, Y. 2000. Subject/Object drop in the Acquisition of Korean: A cross-linguistic

comparison. Journal of East Asian Linguistics 9. 325 – 351. Netherlands: Kluwer

Academic Publishers.

Klima, E. & U. Bellugi. 1966. Syntactic regularities in the speech of children. In J. Lyons &

R. Wales (eds.) Psycholinguistic Papers. Edinburgh: Edinburgh University Press. 183 –

208.

Köpcke, K-M. 1998. The Acquisition of Plural Marking in English and German Revisited.

Journal of Child Language, 25, 293 – 319.

Kuczaj, S. A. & M. P. Maratsos. 1975. What children can say before they will. Merrill-

Palmer Quarterly 21, 89-111.

Labelle, M. 1990. Predication, wh-Movement, and the Development of Relative Clauses.

Language Acquisition 1, 95-119.

Labov, W. 1972. Some principles of linguistic methodology. Language in Society 1(1): 97–

120.

Lacoste, V. 2012. Phonological Variation in Rural Jamaican Schools. Creole Language

Library 42. John Benjamins. Originally published as Doctoral Thesis: University of

Essex, 2008.

310

Lamiroy, B. 1987. The complementation of aspectual verbs in French. Language 63: 278-

298.

Lebeaux, D. 1987. Comments on Hyams. In T. Roeper & E. Williams (eds.), Parameter

Setting. Dordrecht: Foris, 23-39.

Lee, C. 2009. The Acquisition of Modality. In C. Lee, G. Simpson & Y. Kim (eds.), The

Handbook of East Asian Psycholinguistics, Vol. III Korean, 187 – 220. Cambridge

University Press.

Leonard, L., Caselli, M.C., and Devescovi, A. 2002. Italian children’s use of noun and verb

morphology during the preschool years. First Language 22, 287-304.

Levow G.A. 1995. Tense and subject position in interrogatives and negatives in child French:

Evidence for and against truncated structures. In C. T. Schütze, J. B. Ganger & K.

Brohier (eds.), Papers on Language Processing and Acquisition, MIT Working Papers in

Linguistics 26:281-304.

Lewis, Y. E. 2010. Literacy in Elementary School in Jamaica: the case of the grade four

literacy test. Phd dissertation, University of Iowa.

Lillo-Martin, D. 1986. Two Kinds of Null Arguments in American Sign Language. Natural

Language and Linguistic Theory 4, 415 - 444.

Lorusso, P. 2007. The Acquisition of Aspect in L1 Italian. In A. Belikova et al. (eds.),

Proceedings of the 2nd Conference on Generative Approaches to Language

Acquisition North America (GALANA) 253 – 264. Somerville, MA: Cascadilla

Proceedings Project.

Lust, B.C. (ed.) 1986. Studies in the Acquisition of Anaphora: Vol 2. Applying the

Constraints. Dordrecht: Reidel.

Macnamara, J. 1972. Cognitive basis of language learning in infants. Psychological Review

79. 1 - 13.

311

Manzini, M. R. and K. Wexler. 1986. Parameters, Binding Theory and Learnability.

Linguistic Enquiry 18, 413-444.

Maratsos, M. P. 1976. The use of definite and indefinite reference in young children.

Cambridge: Cambridge University Press.

Marrero, V. and Aguirre, C. 2003. Plural acquisition and development in Spanish. In

Montrul, S. and Ordoñez, F. (eds.), Theoretical Linguistics and Language

Development in Hispanic Languages, Sommerville MA: Cascadilla Press. 275 - 296.

Matthewson, L. 2006. Temporal semantics in a supposedly tenseless language. Linguistics

and Philosophy 29: 673 – 713.

McCain, L. 1996. Caribbean Studies Thesis. University of the West Indies, Mona.

Unpublished Undergraduate Thesis.

Meade, R. 2001. Acquisition of Jamaican Phonology. The Netherlands: HIL.

Miller, J.F. 1981. Eliciting procedures for language. In J.F. Miller (ed.), Accessing language

production in children. London: Edward Arnold.

Miller, J. F. & R. S. Chapmann. 1981. The relation between age and mean length of utterance

in morphemes. Journal of Speech and Hearing Research 24, 2, 154 - 161.

Mufwene, S. 2000. Creolization is a social, not a structural, process. In I. Neumann-Holzchuh

and E. Schneider, (eds.), Degrees of restructuring in creole languages, Amsterdam: John

Benjamins, 65 - 84.

Mufwene, S. 2001. The Ecology of Language. Cambridge: CUP.

Muysken, P. & N. Smith. 1990. Question words in pidgin and creole languages. Linguistics

28. 883 – 903. Walter de Gruyter.

Nederstigt, U. 2001. The acquisition of additive “Focus Particles” in German. Proceedings of

the 25th Annual Boston Conference of Language Development. Sommerville,

Cascadilla Press, 554-565.

312

Nelson, K. 1973. Some evidence for the cognitive primacy of categorization and its functional

basis. Merrill-Palmer Quarterly 19. 21 – 39.

Nissenbaum, J. 2000. Investigations of covert phrase movement. Doctoral dissertation, MIT.

Cambridge, MA: MITWPL.

O’ Grady, W. 1999. Towards a new nativism. Studies in Second Language Acquisition, 21.

621 – 633.

Orfitelli, R. M. & N. Hyams. 2008. An experimental study of children’s comprehension of

null subjects: Implications for grammatical/performance account. In H. Chan, H. Jacob, &

E. Kapia, (eds.), Proceedings of the32nd Annual BUCLD, volume 2. 335–346.

Somerville: Cascadilla Press.

Painter, C. 1984. Into the mother tongue: A case study in early language development.

London: Frances Pinter.

Panitsa, G. 2010. Aspects of Aspect: The acquisition of viewpoint and situation aspect in

Modern Greek. Doctoral thesis, University College London.

Papafragou, A. 1998. The Acquisition of Modality: Implications for Theories of Semantic

Representation. Mind and Language Vol. 13 No. 3, 370-399. Blackwell Publishers.

Patrick P. 1999. Urban Jamaican Creole: Variation in the mesolect. Amsterdam: John

Benjamins.

Patrick, P. 2004. Jamaican Creole Morphology and Syntax. In B. Kortmann et al (eds.), A

Handbook of Varieties of English, Vol 2: Morphology and Syntax. Berlin: Mouton de

Gruyter, 407-438.

Patrick, P. 2007. JC Jamaican Patwa (Creole English). Creolica.

Pea, R., R. Mawby & S. MacCain. 1982. World-making and World-revealing: Sematics and

Pragmatics of Modal Auxilliary verbs during the third year of life. Paper presented at

the 7th Annual Boston Conference on Child Language Development.

313

Penner, Z. & T. Roeper. 2000. Trigger theory and the acquisition of complement idioms. In:

N. Dittmar and Z. Penner (eds.), Issues in the Theory of Language Acquisition. Bern:

Peter Lang, 77 - 111.

Perkins, M. 1983. Modal Expressions in English. Frances Pinter: London.

Petronio, K. & D. Lillo-Martin. 1997. Wh-Movement and the position of Spec-CP: Evidence

from American Sign Language. Language, Vol. 73, No.1. 18 – 57.

Piaget, J. 1980. The psychogenesis of knowledge and its epistemological significance. In M.

Piattelli-Palmarini (ed.), Language and learning: The debate between Jean Piaget and

Noam Chomsky. London: Routledge and Kegan Paul.

Pierce, A. 1992. Language Acquisition and Syntactic Theory: A comparative analysis of

French and English child grammars. Dordrecht: Kluwer.

Pinker, S. 1984. Language learnability and language development. Harvard University Press,

Cambridge, MA.

Pizzuto, E. & M. C. Caselli. 1994. The acquisition of Italian verb morphology in a cross-

linguistic perspective. In Y. Levi (ed.), Other children, other languages: issues in the

theory of language acquisition. Hillsdale, NJ: Lawrence Erlbaum. 137 – 187.

Platzack, C. & G. Josefsson. 2000. Subject Omission and Tense in Early Swedish Child

Language. In Working Papers in Scandinavian Syntax 66. 85 – 102.

Pochard, J-C. & H. Devomish. 1986. Deixis in Caribbean English-Lexicon Creole: A

description of A, Da and De. Lingua 69. 105 – 120. North-Holland.

Poeppel D. & K. Wexler. 1993. The full competence hypothesis of clause structure in Early

German. Language 69.1, 1 – 33.

Pollock, J-Y. 1989. Verb Movement, Universal Grammar, and the Structure of IP. Linguistic

Inquiry, Vol. 20, No. 3. 365 – 424. The MIT Press.

314

Pratas, F. & N. Hyams. 2009. Introduction to the Acquisition of Finiteness in Capeverdean. In

Costa, Castro, Lobo & Pratas (eds.), Language Acquisition and Development -

Proceedings of GALANA 2009. Cambridge: Cambridge Scholars Press, pp. 378-390.

Prevost, P. 2009. The Acquisition of French: The development of Inflectional morphology

and syntax in L1 acquisition, bilingualism, and L2 acquisition. Language Acquisition

and Language Disorders 51. Amsterdam/Philadelphia: John Bemjamins.

Pye, C. 1992. The Acquisition of K’iche’ (Maya). In D. I. Slobin (ed.), The Cross linguistic

Study of Language Acquisition, Vol. 3. Erlbaum, Hillsdale, NJ.

Radford, A. 1990. Syntactic Theory and the Acquisition of English Syntax, Oxford:

Blackwell.

Radford, A. 1995. Phrase Structure and Functional Categories. In P. Fletcher & B.

McWhinney (eds.), The Hnadbook of Child Language, Blackwell, Oxford. 483 - 507.

Radford, A. 1996. Towards a structure building model of acquisition. In. H. Clahsen (ed.),

Generative perspectives on language acquisition. Benjamins, Amsterdam. 43 – 89.

Rasetti, L. 2000. Interpretive and Formal Properties of Null Subjects in Early French. In

Generative Grammar in Geneva: 1. 241 – 274.

Rasetti, L. 2003. Optional categories in early French Syntax: A developmental study of root

infinitives and null arguments. Doctoral Thesis: University of Geneva.

Richie, W. C. & T. K. Bhatia. 1999. Child language acquisition: Introduction, foundations,

and overview. In W. C. Ritchie & T. K. Bhatia (eds.), Handbook of language acquisition.

San Diego: Academic Press.

Rizzi L. 1992. Early null subjects and root null subjects. In Geneva Generative Papers 0.1-

2:102-114, republished in B. Lust, G. Hermon & J. Kornfilt (eds.), Binding,

Dependencies and Learnability, vol. 2, Lawrence Earlbaum Associates, Hillsdale, 1994,

249-272. And as chapter 11 of Rizzi (2000).

315

Rizzi, L. 1993/1994. Some notes on linguistic theory and language development: the case of

root infinitives. Language Acquisition 3(4), 371-393.

Rizzi, L. 1997. The fine structure of the left periphery. In L. Haegeman (ed.), Elements of

Grammar. 281 – 337. Dordrecht: Kluwer; also in Rizzi 2000a.

Rizzi, L. 2000. Comparative Syntax and Language Acquisition, Routledge, New York.

Rizzi, L. 2002. On the Grammatical Basis of Language Development: A Case Study.

University of Siena.

Rizzi, L. 2004. Locality and left periphery. In A. Belletti (ed.), Structures and beyond: The

Cartography of Syntactic Structures. Vol. 3. New York: Oxford University Press. 223 –

251.

Rizzi, L. 2005. Phase Theory and the Privilege of the Root. In H. Broekhuis, N. Corver, R.

Huybregts, U. Kleinherz & J. Koster (eds.), Organizing Grammar - Studies in Honor of

Henk van Riemsdijk, Mouton de Gruyter, Berlin, 529 - 537.

Rizzi, L. 2005a. Grammatically-based target-inconsistencies in child language. In K.U. Deen,

J. Nomura, B. Schulz & B.D. Schwartz (eds.), The Proceedings of the Inaugural

Conference on Generative Approaches to Language Acquisition - North America

(GALANA), 19 - 49. Cambridge MA: The MIT Press.

Rizzi, L. & U. Shlonsky. 2007. Strategies of Subject Extraction. In H-M. Gärtner and U.

Sauerland (eds.), Interfaces + Recursion = Language? Chomsky’s Minimalism and the

View from Syntax-Semantics. 115-160. Berlin: Mouton de Gruyter.

Roberts, I. 2007. Diachronic Syntax. Oxford: Oxford University Press.

Roeper, T. 1987. Implicit arguments and the head-complement relation. Linguistic Inquiry

18:267–310.

316

Roeper, T. 1996. The role of Merger Theory and Formal Features in Acquisition. In H.

Clahsen & R. Hawkins (eds.), Generative Perspectives on Language Acquisition,

Emperical Findings, Theoretical Considerations, Cross-linguistic comparisons.

Roeper, T. & de Villiers, J. 1992. The one feature hypothesis for acquisition. Ms., University

of Massachusetts.

Roeper, T. & B. Rohrbacher. 1994. Null Subjects in Early Child English and the Theory of

Economy of Projection. Ms., University of Massachusetts at Amherst.

Roeper, T. & B. Rohrbacher. 2000. True Pro-drop in child English and the principle of

economy or projection. In S. Powers and C. Hamann (eds.), The acquisition of

scrambling and cliticization. 345–396. Dordrecht: Kluwer.

Roper, T. 2007. The Prism of Grammar. Cambridge, MA: MIT Press.

Roper, T. & E. Williams (eds.), 1987 Parameter Setting. Dordrecht: Reidel.

Sano T. & N. Hyams. 1994. Agreement, finiteness, and the development of null arguments.

Proceedings of NELS, 24, 543-558.

Santelmann, L. 1997. Wh-question formation in early Swedish: an argument for continuity,

CP and operators. In Yamakoshi, Someshakar, Blume & Foley (eds.), Cornell Working

Papers in Linguistics.

Sauermann, A., B. Hohle, A. Chen & J. Jarvikivi. 2011. Intonational Marking of Focus in

Different Word Orders in German Children. In M. B. Washburn et al. (eds.), Proceedings

of the 28th West Coast Conference on Formal Linguistics. Somerville, MA: Cascadilla

Proceedings Project.

Savic, S. 1975. Aspects of adult-child communication: the problem of question acquisition.

Journal of Child Language 2. 251 – 260.

Sebba, M. 1987. The syntax of serial verbs: An investigation into serialization in Sranan and

other languages. Amsterdam: John Benjamins.

317

Shirai, Y. & R. Andersen. 1995. The Acquisition of Tense-Aspect Morphology: A Prototype

Account. Language 71, Vol 4. 743 – 762.

Shlonsky, U. 2010. The Cartographic Enterprise in Syntax. Language and Linguistics

Compass 4/6, 417 – 429.

Skinner, B. F. 1948. Verbal Behavior. William James Lectures. Harvard University.

Reproduced by D.C. Palmer 2009.

Slobin, D. I. 1982. Universal and particular in the acquisition of language. In E. Wanner & L.

Gleitman (eds.), Language Acquisition: The state of the art. Cambridge University Press.

Slobin, D. I. 1984. Cross-linguistic evidence for the language-making capacity. In D.I. Slobin

(ed.), The cross-linguistic study of language acquisition. Lawrence Erlbaum Associates.

Stephany, U. 1979. Modality. In P. Fletcher and M. Garman (eds.), Language Acquisition.

Cambridge University Press, 2nd ed. 1986. 375 - 400.

Stewart, M. 2006. Quantification in Jamaican Creole. The Syntax and Semantics of Evri

(‘Every’) in interaction with Indefinites. Doctoral Thesis: University of the West Indies.

Svenonius, P. & C. Kennedy. 2006. Northern Norwegian Degree Questions and the Syntax of

Measurement. In Mara Frascarelli (ed.), Phases of Interpretation, 129 - 157. Berlin:

Mouton de Gruyter.

Syea, A. 1993. Null Subject in Mauritian Creole and the ProDrop Parameter. In Byrne, F. &

J. Holm (eds.), Atlantic Meets Pacific: A Global View of Oidginization and

Creolization. Amsterdam: John Benjamins. 91 - 102.

Szagun, G. 2001. Learning different regularities: The acquisition of noun plurals by German-

speaking children. First Language 21, 109 - 141.

The Language Competence Survey of Jamaica: Data Analysis. Prepared by the The Jamaican

Language Unit, UWI Mona, 2007.

318

Thornton, R. & G. Tesan. 2013. Sentential negation in early child English. Journal of

Linguistics, 49. 367 – 411.

Tomasello, M. 2000. Do young children have adult syntactic competence? Cognition, 74, 209

- 253.

Tomasello, M. 2003. Constructing a Language. A Usage-based theory of language

acquisition, Cambridge, MA: Harvard University Press.

Torrence, H. 2012. The morpho-syntax of silent wh-expressions in Wolof. Natural Language

and Linguistic Theory 30: 1147 – 1184.

Vainikka, A. 1993/4. Case in the development of English Syntax. Langauge Acquisition 3.

257 – 325.

Valian, V. 1990. Null Subjects: a problem for parameter-setting models of language

acquisition. Cognition 35: 105 – 122.

Valian, V. 1991. Syntactic subjects in the early speech of American and Italian children.

Cognition 40:21-81.

Van de Vate, M. 2011. Tense, Aspect and Modality in a Radical Creole: The Case of

Saamaka. Doctoral Thesis: University of Tromso.

Van Valin, R. Jr. 1998. The Acquisition of Wh-Questions and the Mechanisms of Language

Acquisition. In Tomasello (ed.), The New Psychology of Language: Cognitive and

Functional Approaches to Language Structure. London: Lawrence Erlbaum Associates.

Van Wijk, J. 2007. The acquisition of the Dutch plural. Utrecht, LOT.

Veenstra, T. 1996. Serial verbs in Saramaccan: predication and creole genesis. Thesis

Universiteit van Amsterdam. HIL Dissertations.

Vogel, I. & E. Raimy. 2002. The acquisition of compound vs. Phrasal stress: The role of

prosodic constituents. Journal of Child Language 29, 225-250.

319

Wang Q., D. Lillo-Martin, C. Best & A. Levitt. 1992. Null subject versus null object: some

evidence from the acquisition of Chinese and English. Language Acquisition 2. 221 -

254.

Weissenborn, J. 1992. Null Subjects in Early Grammars: Implications for Parameter-setting

Theories. In J. Weissenborn, H. Goodluck and T. Roper (eds.), Theoretical Issues in

Language Acquisition. New Jersey: Lawrence Erlbaum Associates.

Weist, R. H., H. Wysocka, K. Witkowska-Stadnik & E. Buczowska. 1984. The defective

tense hypothesis: On the emergence of tense and aspect in child polish. Journal of Child

Language, 11. 347 - 374.

Wells, G. 1979. Learning and using the auxiliary verb in English. In V. Lee (ed.), Cognitive

Development: Language and thinking from birth to adolescence. London: Croom

Helm, 250-270.

Wells, G. 1985. Language Development in the Pre-school years. Cambridge University Press.

Werner, T. 2003. Deducing the future and distinguishing the past: Temporal interpretation in

modal sentences in English. Ph.D. thesis, Rutgers University.

Westergaard, Marit. 2009. Microvariation as diachrony: A view from acquisition. Journal of

Comparative Germanic Linguistics 12: 49–79.

Wexler, K. 1994. Optional Infinitives, Head Movement, and the Economy of Derivation in

Child Language. In D. Lightfoot & N. Hornstein (eds.), Verb Movement. Cambridge:

Cambridge University Press. 305 - 350.

Wexler, K. 1998. Very early parameter settings and the unique checking constraint: Anew

explanation of the optimal infinitive stage. Lingua 106. 23 – 79.

Wexler, K. 2000. Three Problems in the Theory of the Optional Infinitive Stage:

Stage/Individual Predicates, Eventive Verbs and Finite Null Subjects. In Billerey &

320

Lillehaugen (eds.), WCCFL 19 Proceedings. 560 - 573. Sommerville MA: Cascadilla

Press.

Wexler, K. 2014. A new theory of null-subjects of finite verbs in young children: information-

structure meets phasal computation. In Becker, Grinstead & Rothman (eds.), Generative

Linguistics and Acquisition: Studies in honor of Nina M. Hyams. 325–356.

Winford, D. 1993. Predication in Caribbean English Creoles. Creole Language Library, 10.

Amsterdam: John Benjamins.

Winford, D. 2008. Atlantic Creole Syntax. In S. Kouwenberg & J. V. Singler (eds.), The

Handbook of Creole Studies. Wiley-Blackwell.

Wode, H. 1975. Some stages in the acquisition of questions by monolingual children. Word,

Special Issue on Child Language. 261 - 310.

Wode, H. 1977. On the systematicity of L1 transfer in L2 acquisition. Proceedings from 1977

Second Language Research Forum (SLRF). 160 – 169. Los Angeles: University of

California.

Yamakoshi, K. 2002. Wh-drop in child languages and adult ASL. Proceedings of Console IX,

217–231.

Yang, C. 2002. Knowledge and Learning in Natural Language. Oxford: Oxford University

Press.

Documents

Thesis - Semantic Scholar...Thesis Reference The acquisition of Jamaican Creole: The emergence and transformation of early syntactic systems DE LISSER, Tamirand Nnena Abstract This