thesis - semantic scholar...thesis reference the acquisition of jamaican creole: the emergence and...
TRANSCRIPT
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
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-
Chapter 1 : Introduction
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.
Chapter 1 : Introduction
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
Chapter 1 : Introduction
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.
Chapter 1 : Introduction
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.
Chapter 1 : Introduction
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
Chapter 2: Theoretical Background and Framework
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.
Chapter 2: Theoretical Background and Framework
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).
Chapter 2: Theoretical Background and Framework
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.
Chapter 2: Theoretical Background and Framework
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:
Chapter 2: Theoretical Background and Framework
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.
Chapter 2: Theoretical Background and Framework
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
Chapter 2: Theoretical Background and Framework
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)
Chapter 2: Theoretical Background and Framework
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.
Chapter 2: Theoretical Background and Framework
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
Chapter 2: Theoretical Background and Framework
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
Chapter 2: Theoretical Background and Framework
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
Chapter 3: Background
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.
Chapter 3: Background
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.
3SG PAST PROG eat 3SG food
c. Im did a nyam im fuud.
3SG PAST PROG eat 3SG food
d. Im did a iit im fuud.
3SG PAST PROG eat 3SG food
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.
Chapter 3: Background
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
Chapter 3: Background
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.”
Chapter 3: Background
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.
Chapter 3: Background
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.”
Chapter 3: Background
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:
Chapter 3: Background
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.”
Chapter 3: Background
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
Chapter 4: Methodology
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.
Chapter 4: Methodology
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
Chapter 4: Methodology
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.
Chapter 4: Methodology
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.
Chapter 4: Methodology
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.
Chapter 4: Methodology
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.
Chapter 4: Methodology
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
Chapter 4: Methodology
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
RECDATE AGE(Y;M,D) UTT WORDS MLU
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
Chapter 4: Methodology
38
Table 4: RJU’s MLU Table 5: TYA’s MLU
RECDATE AGE(Y;M,D) UTT WORDS 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
RECDATE AGE(Y;M,D) UTT WORDS MLU
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
Chapter 4: Methodology
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
RECDATE AGE(Y;M,D) UTT WORDS MLU
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
RECDATE AGE(Y;M,D) UTT WORDS MLU
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
Chapter 4: Methodology
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
Chapter 4: Methodology
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
Chapter 4: Methodology
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.
Chapter 4: Methodology
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.
Chapter 4: Methodology
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.
Chapter 4: Methodology
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
Chapter 4: Methodology
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
Chapter 5: General Developmental Patterns in JC
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:
Chapter 5: General Developmental Patterns in JC
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.
Chapter 5: General Developmental Patterns in 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.
Chapter 5: General Developmental Patterns in JC
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:
Chapter 5: General Developmental Patterns in JC
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
Chapter 5: General Developmental Patterns in JC
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
Chapter 5: General Developmental Patterns in JC
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?
Chapter 5: General Developmental Patterns in JC
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.
Chapter 5: General Developmental Patterns in JC
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
Chapter 5: General Developmental Patterns in JC
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
Chapter 5: General Developmental Patterns in JC
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
Chapter 5: General Developmental Patterns in JC
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.
Chapter 5: General Developmental Patterns in JC
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.
Chapter 5: General Developmental Patterns in JC
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.
Chapter 5: General Developmental Patterns in JC
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)
Chapter 5: General Developmental Patterns in JC
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.
Chapter 5: General Developmental Patterns in JC
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.”
Chapter 5: General Developmental Patterns in JC
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.”
Chapter 5: General Developmental Patterns in JC
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.”
Chapter 5: General Developmental Patterns in JC
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.”
Chapter 5: General Developmental Patterns in JC
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
Chapter 5: General Developmental Patterns in JC
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.
Chapter 5: General Developmental Patterns in JC
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)
Chapter 5: General Developmental Patterns in JC
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.
Chapter 5: General Developmental Patterns in JC
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.
Chapter 5: General Developmental Patterns in JC
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
Chapter 5: General Developmental Patterns in JC
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,
Chapter 5: General Developmental Patterns in JC
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.’
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
AGE(Y;M,D) ABL OBL NEC EPIS WI
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
83
Table 3: RJU Modal Production Table 4: TYA Modal Production
RJU
AGE(Y;M,D) ABL OBL NEC EPIS WI
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
AGE(Y;M,D) ABL OBL NEC EPIS WI
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
84
SHU
AGE(Y;M,D) ABL OBL NEC EPIS WI
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
AGE(Y;M,D) ABL OBL NEC EPIS WI
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.”
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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!”
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.”
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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]
Chapter 6: The Acquisition of Tense, Modal and Aspect
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:
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.”
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.”
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
AGE(Y;M,D) COMP PROG PROS RETRO
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
104
TYA
AGE(Y;M,D) COMP PROG PROS RETRO
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
AGE(Y;M,D) COMP PROG PROS RETRO
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
105
SHU
AGE(Y;M,D) COMP PROG PROS RETRO
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
AGE(Y;M,D) COMP PROG PROS RETRO
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and 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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.”
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and 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)
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.”
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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)
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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).
Chapter 6: The Acquisition of Tense, Modal and Aspect
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)
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
INFORMANT COL ALA RJU TYA KEM SHU TOTAL
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
Table 27: TMA Phase 2
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
INFORMANT COL ALA RJU TYA KEM SHU TOTAL
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
Table 28: TMA Phase 3
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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).
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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)
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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)
Chapter 6: The Acquisition of Tense, Modal and Aspect
138
b) Aspprogressive (1;11,4) > Aspprospective (1;11,26) > Modability/permission (2;0,30) > Modwi
(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)
b) Aspprogressive (2;0,21) > Aspprospective (2;6,4) > Modability/permission (2;7,5) > Modwi
(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)
Chapter 6: The Acquisition of Tense, Modal and Aspect
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
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 6: The Acquisition of Tense, Modal and Aspect
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.
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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.
Chapter 7: Null Subject Phenomenon
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.
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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).”
Chapter 7: Null Subject Phenomenon
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.
Chapter 7: Null Subject Phenomenon
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.”
Chapter 7: Null Subject Phenomenon
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.
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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
AGE (Y;M,D) UTTERANCES
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
Chapter 7: Null Subject Phenomenon
160
AGE (Y;M,D) UTTERANCES
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
Chapter 7: Null Subject Phenomenon
161
AGE (Y;M,D) UTTERANCES
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
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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-
Chapter 7: Null Subject Phenomenon
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?”
Chapter 7: Null Subject Phenomenon
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?”
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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.
Chapter 7: Null Subject Phenomenon
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”
Chapter 7: Null Subject Phenomenon
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?”
Chapter 7: Null Subject Phenomenon
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
“Where are you 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)
Chapter 7: Null Subject Phenomenon
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.
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 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
Chapter 7: Null Subject Phenomenon
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.
Chapter 7: Null Subject Phenomenon
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)
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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?”
Chapter 7: Null Subject Phenomenon
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.
Chapter 7: Null Subject Phenomenon
178
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 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
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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)
Chapter 7: Null Subject Phenomenon
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”
Chapter 7: Null Subject Phenomenon
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.
Chapter 7: Null Subject Phenomenon
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.
Chapter 7: Null Subject Phenomenon
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?”
Chapter 7: Null Subject Phenomenon
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.
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 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
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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.
Chapter 7: Null Subject Phenomenon
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.
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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
Chapter 7: Null Subject Phenomenon
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.
Chapter 7: Null Subject Phenomenon
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%
Chapter 7: Null Subject Phenomenon
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.
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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)
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.”
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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”
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
203
“Where is the ball?”
10) Wei yaa go ti? (TYA 2;06)
where 2SG~PROG go
“Where are you going?”
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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:
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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?”
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.”
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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”
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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).
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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).
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.”
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.”
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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).
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.”
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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)
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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).
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.”
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.”
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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-
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.
Chapter 8: The Acquisition of Wh-Interrogation, Focalization and Topicalization
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.
Chapter 9: The Acquisition of Typical Creole Features
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.”
Chapter 9: The Acquisition of Typical Creole Features
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)
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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.”
Chapter 9: The Acquisition of Typical Creole Features
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.
Chapter 9: The Acquisition of Typical Creole Features
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.”
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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.
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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.”
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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.
Chapter 9: The Acquisition of Typical Creole Features
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):
Chapter 9: The Acquisition of Typical Creole Features
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)
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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.”
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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’.
Chapter 9: The Acquisition of Typical Creole Features
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)
Chapter 9: The Acquisition of Typical Creole Features
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.”
Chapter 9: The Acquisition of Typical Creole Features
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:
Chapter 9: The Acquisition of Typical Creole Features
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.
MLU COL ALA RJU TYA KEM SHU TOTAL
< 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
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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.”
Chapter 9: The Acquisition of Typical Creole Features
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.
MLU COL ALA RJU TYA KEM SHU TOTAL
< 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)
Chapter 9: The Acquisition of Typical Creole Features
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.”
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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.”
Chapter 9: The Acquisition of Typical Creole Features
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.
Chapter 9: The Acquisition of Typical Creole Features
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.
Chapter 9: The Acquisition of Typical Creole Features
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):
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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.”
Chapter 9: The Acquisition of Typical Creole Features
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.”
Chapter 9: The Acquisition of Typical Creole Features
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.
MLU COL ALA RJU TYA KEM SHU TOTAL
< 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:
Chapter 9: The Acquisition of Typical Creole Features
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.
Chapter 9: The Acquisition of Typical Creole Features
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)
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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)
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 9: The Acquisition of Typical Creole Features
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
Chapter 10: Conclusions and Implications
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
Chapter 10: Conclusions and Implications
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
Chapter 10: Conclusions and Implications
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
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 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
INFORMANT COL ALA RJU TYA KEM SHU TOTAL
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
INFORMANT COL ALA RJU TYA KEM SHU TOTAL
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.