Improving Early Literacy Outcomes : Curriculum, Teaching, and Assessment [1 ed.] 9789004402379, 9789004402355

Citation preview

Improving Early Literacy Outcomes: Curriculum, Teaching, and Assessment

IBE on Curriculum, Learning, and Assessment Series Editor Mmantsetsa Marope (IBE UNESCO, Switzerland) Managing Editor Simona Popa (IBE UNESCO, Switzerland) Editorial Board Manzoor Ahmed (BRAC University, Bangladesh) Ivor Goodson (University of Brighton, UK) Silvina Gvirtz (Universidad de San Andrés, Argentina) Hugh McLean (Open Society Foundations, UK) Natasha Ridge (Al Qasimi Foundation for Policy Research, UAE) Joel Samoff (Stanford University, USA) Yusuf Sayed (Cape Peninsula University of Technology, South Africa) Nelly Stromquist (University of Maryland, USA) Felisa Tibbitts (Teachers College, Columbia University, USA) N. V. Varghese (National University of Educational Planning and Administration, India)

VOLUME 4

The titles published in this series are listed at brill.com/ibe

Improving Early Literacy Outcomes Curriculum, Teaching, and Assessment Edited by

Nic Spaull and John P. Comings

අൾංൽൾඇ_ൻඈඌඍඈඇ

All chapters in this book have undergone peer review. The Library of Congress Cataloging-in-Publication Data is available online at http://catalog.loc.gov

ISBN 978-90-04-40233-1 (paperback) ISBN 978-90-04-40235-5 (hardback) ISBN 978-90-04-40237-9 (e-book) Copyright 2019 jointly by IBE-UNESCO and Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Brill Hes & De Graaf, Brill Nijhoff, Brill Rodopi, Brill Sense, Hotei Publishing, mentis Verlag, Verlag Ferdinand Schöningh and Wilhelm Fink Verlag. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change. This book is printed on acid-free paper and produced in a sustainable manner.

CONTENTS

Foreword Mmantsetsa Marope

vii

Notes on the Chapters

xi

Notes on Contributors

xv

Introduction: Reading and Writing for Meaning and Pleasure Nic Spaull

1

1.

The Early Reading Curriculum: International Policy and Practice Claire McLachlan

11

2.

Aligning Curriculum and Assessment in Early Reading Education Peter Afflerbach

29

3.

Assessing Early Literacy Outcomes in Burkina Faso and Senegal: Using DHS and PASEC to Combine Access and Quality Nic Spaull and Adaiah Lilenstein

47

Getting It Right from the Start: Some Cautionary Notes for Early Reading Instruction in African Languages Elizabeth J. Pretorius

63

The Teaching of Reading and Writing in Second- and Multi-Language Contexts: Evidence from Canada Robert Savage and Marie-France Côté

81

4.

5.

6.

(DUO\/LWHUDF\,QVWUXFWLRQLQ,QGLD5HGH¿QLQJWKH&KDOOHQJH Shobha Sinha

7.

Challenges Associated with Reading Acquisition in Sub-Saharan Africa: Promotion of Literacy in Multilingual Contexts Heikki Lyytinen, Emma Ojanen, Jacqueline Jere-Folotiya, Stella Damaris Ngorosho, Francis Sampa, Pamela February, Flora Malasi, Jonathan Munachaka, Christopher Yalukanda, Kenneth Pugh, and Robert Serpell

8.

Entering into the Written Culture to Overcome Inequalities: Teaching Literacy to Children from Vulnerable Communities Alejandra Medina

v



119

133

CONTENTS

9.

Those Children Who Are Left Behind: A Reading Programme That Works Beatriz Diuk

155

10. We Want to Learn: A Programme for the Linguistic, Cognitive, and Socio-Emotional Development of Young Children in Argentina Ana María Borzone and Mariela Vanesa De Mier

171

11. Powerful Reforms in Early Language and Literacy Instruction in India Shailaja Menon, Sajitha S. Kutty, Neela Apte, Abha Basargekar, and Ramchandar Krishnamurthy 12. Integrating Curriculum, Teaching, and Learning Materials to Improve Learning Outcomes in Early Grade Reading: A Common Goal for Burkina Faso, Niger, and Senegal Amapola Alama 13. Teaching and Learning to Read and Write in a Multilingual Context: Burkina Faso, Niger, and Senegal Bernard Schneuwly, Sandrine Aeby Daghé, Irina Leopoldoff, Glaís Sales Cordeiro, Thérèse Thévenaz-Christen, and Simon Toulou 14. Teaching Reading in Burkina Faso: Improving Reading and Writing Outcomes in the Early Years of Primary School Sandrine Aeby Daghé, Irina Leopoldoff, Thérèse Thévenaz-Christen, and Victor Yaméogo 15. Modules to Train Teachers to Teach Reading and Writing in Niger: From an Analysis of the Current Situation to a Collaborative Production of Tools for Teacher Education Bernard Schneuwly, Simon Toulou, and Maman Mallam Garba 16. Senegal: Improving Training to Bolster Reading and Writing Skills in French as a Second Language Thérèse Thévenaz-Christen and Glaís Sales Cordeiro

vi

191

207

229

245

263

279

FOREWORD

The breathtakingly rapid pace of change in the 21st century amplifies the pertinence of education and learning systems as foundations and key sources of lifelong learning and of human resilience. The emergence of the fourth industrial revolution is pressuring learners to develop a wider range of multifaceted, multidisciplinary, complex, and integrated competences, yet all of these are built on the foundational competence of reading for meaning. While policies about the role of education in development are commonplace, specific and concrete instruments for enacting these policies remain both scant and ineffective. In today’s world, the heightened role of education in human and societal development coexists with heightened frustration about the inability of schooling to properly facilitate learning and to prepare millions of students for the future. The world had counted on the formal education systems to fulfil their core mandate of facilitating learning and, thus, to meet the Internationally Agreed Goals (IAGs) for learning. The Education for All (EFA) movement enabled an unprecedented expansion of access to formal education, mainly through schooling. However, for many learners, enrolment did not translate into learning (UNESCO, 2014). Given that more than half of all children are not acquiring basic competences worldwide (UIS, 2017a), it was crucial for a global post-2015 goal to be set, to monitor whether all children and youth, regardless of their circumstances, acquire foundation skills in reading, writing, and mathematics by 2030. Governments committed to ensure an “inclusive and equitable quality education and the promotion of lifelong learning opportunities for all”, as encompassed in the UN Sustainable Development Goal (SDG) 4. Yet, two years into the monitoring of SDG 4.1, which requires primary and secondary education that leads to “relevant and effective learning outcomes”, the UIS reported that 617 million children and adolescents worldwide were not achieving minimum proficiency levels in reading and mathematics. This means that more than half – 56% – of all children won’t be able to read or handle mathematics with proficiency by the time they are of age to complete primary education. The proportion is even higher for adolescents, with 61% unable to achieve minimum proficiency levels when they should be completing lower secondary school (UIS, 2017b). This signals a global learning crisis that persists despite the increasing attention given to learning. The education community needs to find effective ways to address this alarming situation, which is holding back millions of children and threatening progress and stability worldwide. Thus, this admirable book, which focuses on improving early literacy outcomes, could hardly be more timely.

vii

FOREWORD

The idea of the book stemmed from the results of a three-year International Bureau of Education (IBE) project, Improving Learning Outcomes in Early Grade Reading: Integration of Curriculum, Teaching, Learning Materials, and Assessment. Funded by the Global Partnership in Education (GPE), and involving a research team from the University of Geneva, the project was conducted in three sub-Saharan African countries: Burkina Faso, Niger, and Senegal. The sharing of vision and policy experiences that took place throughout that project was instrumental in developing an informed outline of the current book. An overview of the project, its theoretical underpinnings, and specific national cases are presented in this volume. Gradually, the book grew thicker and richer, including an overview of international policy and practice in the early reading curriculum and a solid theoretical chapter on aligning curriculum and assessment in early reading education. The case of dual and multi-language education in Canada came as a natural addition, as it refers to language immersion in a developed country, which is a policy option in developing ones. Cases of Argentina, Chile, and India followed. With the addition of important work on early reading instruction in African languages and on assessing early literacy outcomes in Burkina Faso and Senegal, the book has become an expanded guide to improving early reading outcomes in developing countries, with a special focus on some of the most understudied countries in the world (e.g., Burkina Faso, Niger, and Senegal). Sixteen interconnected chapters cast an ever-vigilant and deflationary eye on the temptation to look at a single undifferentiated goal for literacy without taking into consideration the crucial value of context. Through probing analyses of research, policy, and practice, the book argues that improving the teaching, learning, and assessment of early grade literacy is key not only to expanding quality, access, and equity of education but also to unlocking all the other SDGs, and ultimately to driving development. Nic Spaull and John P. Comings, leading experts in the field of literacy, edited the book with intellectual gravitas and enthusiasm. I am grateful for their excellent work. I would also like to thank Amapola Alama and her team, including Aurélie Boulos, Annabel Conway, Ryo Kato, Fiona Moreno, Julia Napoli, and Louisa Rowe, who contributed at different stages of this project. Improving Early Literacy Outcomes is unlike many other books about literacy today: it carries the weight of scientific authority, while being rooted in different yet common experiences of high aspirations repeatedly confronting harsh realities. It dives into uncharted territory, by featuring countries where there is a significant lack of research and theory on early reading. It also features research done by scholars who live and work in those countries and regions, and who bring a wealth of contextual insights to their analyses. As engaging as it is accessible, the book is an indispensable guide to policymakers, practitioners, educators, academics, and anyone interested in improving early grade reading.

viii

FOREWORD

REFERENCES UIS [UNESCO Institute for Statistics] (2017a). More than one-half of children and adolescents are not learning worldwide. Fact sheet 46. Montreal: UIS. UIS (2017b). Reducing global poverty through universal primary and secondary education. Fact sheet 44. Montreal: UIS. UNESCO (2014). Teaching and learning: Achieving quality for all. EFA global monitoring report. Paris: UNESCO.

Mmantsetsa Marope, Director UNESCO International Bureau of Education (IBE) Geneva, Switzerland

ix

NOTES ON THE CHAPTERS

Introduction. Nic Spaull provides the conceptual and empirical framing for the book, arguing that reading and writing in developing countries remain underresearched and under-theorized. By reviewing the “5 T’s of literacy” (Teaching, Texts, Tests, Training, and Tongue), he shows how the current book contributes to our understanding of how literacy unfolds in developing countries. The chapter also provides an overview of reading outcomes in developing countries, and in )UDQFRSKRQH:HVW$IULFDVSHFL¿FDOO\ Chapter 1. Claire McLachlan makes a case for a comprehensive approach to early-grade reading instruction that is based on research into how children acquire and improve reading skills and how children are motivated to use reading for pleasure, learning, and interaction within their culture and with the wider world culture. She also makes a case for adapting the practices developed in the cultures of OECD countries to the cultures in African countries to ensure the effectiveness of these imported teaching and learning programmes. Chapter 2. Peter Afflerbach sets out a four-part process: (1) define literacy; (2) identify the skills and knowledge needed to acquire literacy; (3) develop a curriculum and instruction that supports students acquiring those skills and knowledge; and (4) design an assessment to measure how well the curriculum and instruction build the skills and knowledge students need. In this chapter, he suggests three factors that are critical to learning the skills and knowledge of literacy: (1) motivation, (2) selfefficacy, and (3) metacognition. He also suggests that measures of these three factors should be added to formative and summative assessments. Chapter 3. Nic Spaull and Adaiah Lilenstein look at educational statistics in Burkina Faso and Senegal and note that both countries have low school completion rates in Grades 2 and 5, as well as low reading scores at these two grade levels. They then combine the measures of grade completion and reading scores to calculate the percentage of children in the Grade 2 age cohort (both those in school and those who have dropped out) who have enough grade-level reading skills. This comprehensive approach provides a more accurate measure of success. They also look at the demographic characteristics of sex and family wealth and how they predict the more comprehensive measure that combines Grade 2 completion and reading skill acquisition. Chapter 4. Elizabeth J. Pretorius describes the differences between some African languages and the European languages that are the subject of most reading research. She presents the research on learning to read in African languages and highlights how this research could inform the teaching of reading in countries that have nonEuropean languages of instruction.

xi

NOTES ON THE CHAPTERS

Chapter 5. Robert Savage and Marie-France Côté explore dual language literacy learning through a case study on Canada. Canada has over 40 years’ experience and research to draw on and, therefore, offers insights into one country that has been successful at bilingual education. The chapter reviews contemporary theory and research relating to dual language instruction and identifies key aspects of effective dual language instruction that could potentially be generalized to other countries. Chapter 6. Shoba Sinha employs an emergent-literacy perspective to describe how children who live in low-literacy homes manage literacy in school, what body of knowledge a teacher must have to teach children who come from low-literacy homes, and the impact on learning of the context of schools where teaching takes place. In addition, she suggests alternative ways of conceptualizing early literacy programmes based on interventions in India and examines the challenges involved in implementing programs of improved literacy instruction. Chapter 7. Heikki Lyytinen and his co-authors note the problems of poorly trained teachers and large class sizes in Africa. Drawing on experience in Zambia, they present a case study of a digital game that can help overcome these problems and lead to more effective literacy instruction, particularly for students facing barriers to learning how to read. Chapter 8. Alejandra Medina describes a project that was conducted with students in primary schools that serve vulnerable groups in San Antonio, Chile. This four-year project combined cultural learning with the teaching of language and literacy. The project design included increasing daily opportunities for reading complex texts, contextualizing the texts, and writing about the texts. This approach led to a statistically significant impact on student reading comprehension and text production. Chapter 9. Beatriz Diuk describes DALE!, a reading program that takes into account three key aspects of the apprenticeship model: (1) becoming a member of a community of practice; (2) distinguishing between exposure to the complete, meaningful process and instances of practicing partial tasks; and (3) organizing learning within a framework of guided participation. Chapter 10. Ana María Borzone and Mariela Vanesa De Mier describe the We Want to Learn programme that serves young children (between five and seven years old) from different social groups in Argentina. The program builds on the foundation of an earlier programme, which focuses on developing oral and written language, and improves on it by adding insights from cognitive psychology and neuroscience, which identify the need to help children develop memory, attention, and executive function skills. Chapter 11. Shailaja Menon and others acknowledge that India’s national literacy rate has gone up significantly since it gained independence in 1947, but make the case that deeply entrenched cultural beliefs and practices surrounding language and literacy, as well as their related curricula and pedagogy, remain unchanged in many Indian classrooms. They then suggest ways in which India could overcome these beliefs and practices. xii

NOTES ON THE CHAPTERS

Chapter 12. Amapola Alama describes a cooperative project between UNESCO’s International Bureau of Education (IBE) and the governments of Burkina Faso, Niger, and Senegal, to improve literacy learning outcomes in the first three years of primary education. Though the project goal was the same for all three countries, the ways in which that goal would be achieved was tailored to each country. This meant considering each country’s unique historical, socio-economic, linguistic, and cultural contexts. The description presents the way in which the IBE works with countries, the conceptual framework that guided this project’s implementation, and the preliminary results of this collaboration. Chapter 13. Bernard Schneuwly and his team explore the theoretical background for, and research conducted within, the first phase of the three-country project described in Chapter 12. The chapter focuses on the integration of curriculum, teaching, learning materials, and assessment within the overall project. After describing the historical, linguistic, and educational context of the three countries, the authors outline the conceptual and theoretical bases for their fieldwork, which are founded upon the essentials of language acquisition and instruction, provide an overview of each country analysis, and describe a framework for the three case studies presented in Chapters 14, 15, and 16. Chapter 14. Sandrine Aeby Daghé and co-authors provide an analysis of the current curricular literature and reading instruction practices in Burkina Faso to identify the adjustments required to bring about a curriculum alignment that will enhance learners’ progress in reading and writing. Chapter 15. Bernard Schneuwly, Simon Toulou, and Maman Mallam Garba provide a similar analysis with regard to Niger. Chapter 16. Thérèse Thévenaz-Christen and Glaís Sales Cordeiro do the same with regard to Senegal.

xiii

NOTES ON CONTRIBUTORS

Sandrine Aeby Daghé (Switzerland) teaches French didactics at the Department of Educational Sciences, University of Geneva. Her research focuses on the teaching and learning of written language (reading and writing) and literature. She is particularly interested in the study of the early learning of written language and of teachers’ gestures in literature class. Peter Afflerbach (United States) is a professor of education at the University of Maryland, College Park, Maryland. He serves on the Standing Reading Committee of the National Assessment of Educational Progress (NAEP), and is the Chair of the Reading Assessment Committee of the International Literacy Association. His research interests include reading assessment, individual differences in reading, reading strategies, and the verbal reporting methodology. Amapola Alama (France) is an education consultant, with over 25 years’ experience supporting ministries of education all over the world in their reform processes. Between 2011 and 2018, she was a senior programme specialist at the UNESCO International Bureau of Education (IBE). Previously, she worked as a programme director at the Ibero-American States Organization for Education, Science, and Culture, based in Spain and in Morocco; and as head of the education sector at the Spanish Development Agency. Neela Apte (India) works at Shantilal Muttha Foundation, in the field of values education. Prior to that, she worked as a researcher in the Literacy Research in Indian Languages (LiRIL) project. Neela’s work experience includes international consultancies to develop first-language workbooks for primary grades; and involvement with School Management Committees, as a part of the Devdatta Dabholkar Fellowship programme. Abha Basargekar (India) is currently pursuing her PhD in Development Psychology at the University of Virginia, prior to which she was a graduate student at the University of Chicago. She worked as a junior project officer in QUEST, a nongovernmental organization of Maharashtra, where she was deeply involved with the work of the longitudinal research project, Literacy Research in Indian Languages (LiRIL), and before that worked at the Rural Education Centre, Rishi Valley. Ana María Borzone (Argentina) is a principal investigator of the National Council of Technical and Scientific Research (CONICET), where she has been working since 1969. She directs research projects on the linguistic and cognitive aspects of literacy in groups at risk for poverty: children, youth, illiterate adults, and ethnic minorities. xv

NOTES ON CONTRIBUTORS

John P. Comings (United States) is a senior technical consultant at World Education, an international NGO based in Boston, and a faculty member at the Center for International Education, University of Massachusetts at Amherst. He is also lead early grade reading senior consultant for All Children Reading: A Grand Challenge for Development. Marie-France Côté (Canada) is professor at Université du Québec à Montréal (UQAM). She is particularly interested in developing interventions that could help prevent reading and writing difficulties, as well as remedial interventions for struggling students. More specifically, most of her research concerns the transfer of learning in relation to literacy development in multilingual as well as monolingual contexts. Stella Damaris Ngorosho (Tanzania) is currently a lecturer of Special Needs Education programmes at Sebastian Kolowa Memorial University (SEKOMU), Tanzania. She gained her doctoral degree at Åbo Akademi University, Finland. Her research focuses on adult literacy, learning, and development; dyslexia prevention through digital games; and the impact of education leadership and management on learning. Mariela Vanesa De Mier (Argentina) is an assistant researcher of the National Council of Technical and Scientific Research (CONICET). She teaches courses on language teaching (undergraduate and postgraduate level) at Faculty of Education, Pontifical Catholic University of Argentina (UCA), and at the National University of Buenos Aires (UBA). Her current research focuses on the role prosody plays in learning to read. Beatriz Diuk (Argentina) is a research scientist in Argentina’s National Council for Science and Technology (CONICET) and a professor at the University of San Martín, where she coordinates the Center for Applied Psychopedagogical Research. Her interest in bridging the gap between research and educational practices led her to construct DALE!, a literacy programme for children who have not been able to acquire even basic reading and spelling skills, despite their attending school. Pamela February (Namibia) teaches Educational Psychology and Inclusive Education at the University of Namibia, Faculty of Education. She received her PhD in Psychology from University of Jyväskylä, Finland. Her research focuses on reading acquisition and fluency; vocabulary acquisition; inclusive education practices; deaf education; and the use of technology to enhance childhood learning. Jacqueline Jere-Folotiya (Zambia) is a lecturer and researcher in the Department of Psychology, University of Zambia, and coordinates the Centre for the Promotion of Literacy in sub-Saharan Africa (CAPOLSA). Her research interests include early xvi

NOTES ON CONTRIBUTORS

childhood development, especially early childhood education, early grade literacy acquisition, teacher training, and assessments in early childhood. Ramchandar Krishnamurthy (India) is associate director of teacher education and student affairs at Azim Premji University, Bangalore, India. His current interests are in quantitative reasoning using computer programmes, reasoning, and argumentation; and in developing sociophilosophical perspectives in school education. Sajitha S. Kutty (India) is a co-founder of SAJAG, a non-governmental organization that works with children and young adults on early literacy and language development in Maharashtra. Prior to that, she worked as a researcher on the longitudinal research project Literacy Research in Indian Languages (LiRIL). She has also worked with other non-governmental agencies in the areas of education and women’s empowerment, most notably, QUEST, Maharashtra; and Samaj Pragati Sahayog, Madhya Pradesh. Irina Leopoldoff (Russian Federation) is a lecturer at the Department of Educational Sciences, University of Geneva. Her research focuses on Vygotsky’s œuvre on pedology, teaching and learning practices of written and spoken expression in the first language classroom, and emotion in learning and teaching literature. Adaiah Lilenstein (South Africa) is a researcher at the Research on Socioeconomic Policy (RESEP) group, Stellenbosch University, South Africa. Her research focuses on economic development, especially relating to education, poverty, inequality, and the labour market. Heikki Lyytinen (Finland) is emeritus professor of developmental neuropsychology, and UNESCO Chair on Inclusive Literacy Learning for All, at the University of Jyväskylä, Finland. His main research-related interest concerns the development of scientifically validated digital learning environments to support globally the acquisition of basic scholastic skills, most specifically reading skills. Flora Malasi (Kenya) is deputy academic registrar, lecturer, and researcher at Kenya Institute of Special Education (KISE) and a PhD student at Maseno University. She chairs the learning disability survey research project called 2019 Kenya, and is a master trainer of the national technical team of the TUSOME early grade literacy programme of the Kenya Ministry of Education. Her research centers on special needs education, assessment, and early intervention. Maman Mallam Garba (Niger) is a professor and a researcher at the University of Niamey and a technical advisor to the minister for primary education and literacy. He worked for the German Technical Cooperation (GTZ), where he contributed to the implementation of education projects and to the production of several xvii

NOTES ON CONTRIBUTORS

publications on bilingual education, reproductive health, teacher training, and linguistic research. Claire McLachlan (Australia) is dean of the School of Education, Federation University in Victoria, Australia. She is also an honorary professor at the Wilf Malcolm Institute for Educational Research at the University of Waikato, New Zealand. She has taught education at Massey University, University of WisconsinMadison, University of Auckland, Auckland University of Technology, and University of Waikato. Claire has published 9 books and has over 200 academic outputs. She has edited the practitioner journal, Early Education, since 2006, and is a regular reviewer for many journals and academic publishers. Alejandra Medina (Chile) is a senior education consultant. For 13 years, she taught reading and writing to first graders at the French School of Santiago. She worked in the Ministry of Education, alongside a team of professionals, to develop literacy instruction and educational resources for Chilean students, training teachers, school supervisors, and scholars. Shailaja Menon (India) is professor and programme-in-charge of the Early Literacy Initiative at Tata Institute of Social Sciences, Hyderabad, Telengana. She also teaches at the School of Education, Azim Premji University, Bangalore. She was a principal investigator of the longitudinal project, Literacy Research in Indian Languages (LiRIL). Shailaja is the founder of an annual children’s literature festival, KathaVana, hosted by Azim Premji University each year in Bangalore, Karnataka. Jonathan Munachaka (Zambia) is a lecturer in the Department of Educational Psychology, Sociology, and Special Education of the University of Zambia. He is currently a PhD student in Psychology at the University of Jyväskylä, Finland. Emma Ojanen (Finland) is a clinical psychologist trained at University of Jyväskylä, Finland. She has concentrated since 2007 on documenting the efficiency of GraphoLearn technology in helping African children to acquire the basic reading skill in Kenya, Namibia, Tanzania, and Zambia. Elizabeth J. Pretorius (South Africa) is a professor in the Department of Linguistics and Modern Languages, University of South Africa (Unisa). Her teaching fields include phonology, text linguistics, psycholinguistics (first-language acquisition), and applied linguistics (additional/second-language teaching and learning). Her main area of research is reading literacy, especially in high poverty contexts. Kenneth Pugh (United States) is president and director of research at Haskins Laboratories, a Yale University and University of Connecticut affiliated interdisciplinary institute, dedicated to the investigation of the biological bases of xviii

NOTES ON CONTRIBUTORS

language. He also holds academic appointments in the Department of Psychology, University of Connecticut; Department of Linguistics, Yale University; and Department of Diagnostic Radiology, Yale University School of Medicine. Glaís Sales Cordeiro (Brazil) is a senior lecturer at the Faculty of Psychology and Educational Sciences, University of Geneva. Her research focuses on oral language, reading, and writing instruction in the first years of primary school. She is the codirector of the Maison des Petits Network and a member of the Research Group for the Analysis of French Teaching (GRAFE). Francis Sampa (Zambia) is a project manager/ team lead, working with USAID/ Zambia Education Data Activity, to provide distinct assessment, data management, and research and evaluation services to monitor and track the progress of USAID’s Let’s Read Zambia. He received his PhD in Psychology, from University of Jyväskylä, Finland. Robert Savage (Canada) is head of the Department for Psychology and Human Development and Professor of Psychology at the Institute of Education, University College London (UCL). His main research interests involve the effects of reading interventions, effective teaching, and basic cognitive processes in literacy. Bernard Schneuwly (Switzerland) is full professor of Didactics of Language at the University of Geneva, Department of Educational Sciences. He conducts research on the history of teaching first languages; didactics as an emerging research field in Europe; teaching methods for oral and written expression; the construction of objects of teaching in ordinary first-language classroom practices; and the relationship between teaching and development within a historic-cultural perspective (Vygotsky). He was the president of the Department of Educational Sciences, dean of the Faculty of Psychology and Educational Sciences, and founder and director of the Institute of Teacher Education at the University of Geneva. Robert Serpell (United Kingdom) is professor at the University of Zambia. From 2003 to 2006, he was vice-chancellor of the University and he is the current chancellor of Eden University. His current research foci include early childhood development, literacy promotion, education for children with special needs, cultural psychology, and applications of psychology to public policy. In 2017, he received the Society for Research in Child Development (SRCD) Award for distinguished contributions to understanding international, cultural, and contextual diversity in child development. Shobha Sinha (India) is an associate professor in the Department of Education, University of Delhi. Her research interests include early literacy of children from low-literacy backgrounds, literacy in classroom contexts, and response to literature. xix

NOTES ON CONTRIBUTORS

Nic Spaull (South Africa) is a senior researcher at the Research on Socioeconomic Policy (RESEP) group in the Economics Department, Stellenbosch University. His research interests center on education policy in South Africa and sub-Saharan Africa, as well as early grade reading and assessment in African languages. He consults for UNESCO, the OECD, the World Bank, the South African Presidency, and advises a number of Trusts and Foundations. He is also the director of the ‘Funda Wande: Reading for Meaning’ project, an open-access multi-media course to teach Foundation Phase teachers (grades R-4) how to teach reading and writing. Thérèse Thévenaz-Christen (Switzerland) conducts research with the Research Group for the Analysis of French Teaching (GRAFE), led by Bernard Schneuwly and Joaquim Dolz. Her work concerns teaching and learning of comprehension and production of spoken and written texts. Simon Toulou (Cameroon) is currently head of practical training courses and field trainers for secondary education at the Teacher Training Institute, University of Geneva. He is a member of the Research Group for the Analysis of French Teaching (GRAFE), led by Bernard Schneuwly and Joaquim Dolz. Christopher Yalukanda (Zambia) is a researcher at the Zambia National Union of Teachers (ZNUT). He is also a doctoral student at University of Jyväskylä, Finland. He was the first chairperson of the Teaching Council of Zambia, a body that registers and regulates the professional practice and conduct of teachers in Zambia. His research interests are in developmental psychology, teacher education, reading instruction, teacher motivation and early childhood education. Victor Yaméogo (Burkina Faso) teaches and is a teacher trainer and coordinator at the École normale supérieure (ENS), University of Koudougou, Burkina Faso. His research focuses on philosophy of education and bilingual education in Burkina Faso.

xx

NIC SPAULL

INTRODUCTION Learning to Read and Write for Meaning and Pleasure

Learning to read and write for meaning and pleasure are arguably the two most important skills that children learn in primary school. Virtually all forms of social, civic, and economic engagement in the modern world are now mediated through text. Using the skill of reading for meaning, young adults are expected to read everything from newspapers and novels to poetry and political manifestos. It is for this reason that the global community has committed itself to ensure that by 2030 all children complete free, quality primary and secondary education that leads to relevant learning outcomes (UNESCO, 2017). Yet, a consistent body of research points to the fact that the vast majority of children in developing countries do not acquire this fundamental skill (Pritchett, 2013; Muralidharan, 2013; Spaull and Taylor, 2015). While almost all children will naturally acquire oral language (the systematic representation of meaning using speech sounds), fewer than half will learn how to read (the systematic encoding of meaning using print). Around the world reading researchers are increasingly turning their attention to early literacy in developing countries and asking why so many children don’t learn to read despite their having access to formal schooling (World Bank, 2018). The current volume brings together researchers from Latin America (Argentina and Chile), India, Southern Africa (South Africa and Zambia), and West Africa (Burkina Faso, Niger, and Senegal), providing diverse perspectives on early literacy in their respective countries. Drawing together quantitative researchers, linguists, and educationists from four continents, the book shows that irrespective of epistemologies or methodologies, all researchers stress the need for a more nuanced understanding of how early literacy unfolds in developing countries. Importantly, the majority of these researchers live and work in the countries and regions they are analysing, bringing contextual insights that may be overlooked by researchers living abroad. READING IN DEVELOPING COUNTRIES: UNDER-RESEARCHED AND UNDER-THEORIZED

Perhaps the most important contribution of this edited volume is that it adds to an almost-non-existent body of literature on early grade reading in some of the most under-studied countries in the world, such as Burkina Faso, Niger, and Senegal (see

© JOINTLY BY IBE-UNESCO AND KONINKLIJKE BRILL NV, LEIDEN, 2019 DOI: 10.1163/9789004402379_001

SPAULL

chapters 3 and 12 through 16). This dearth of research on early grade reading also extends, however, throughout Asia and into the rest of Africa. Languages spoken and used in the developing world remain under-researched and under-theorized. This is true not only of minority languages in so-called ‘small’ countries or communities, but also of such widely spoken and widely read languages as Arabic and Hindi. A cursory bibliometric analysis of three leading international journals of reading research1 showed that 68% of the 1,423 articles published in the last 10 years (2009– 2019) focused on reading in either English or French. The combined number of articles published on reading in Arabic, Hindi, Bengali, and Swahili amounted to a mere 6% of the total. These four languages alone cover over a billion people in the developing world (Ethnologue, 2019). This lack of research and theory is problematic when one considers the current scholarly consensus that children should first learn to read in a language they speak and understand (Snow, 2017; August and Shanahan, 2006). Emerging quantitative research from South Africa and Kenya points to the clear benefits of mother-tongue instruction in the first years of school (Taylor and von Fintel, 2016; Piper et al., 2016; but see also Piper et al., 2018, for a cautionary note). To illustrate the centrality of language in early grade reading – and to provide a helpful heuristic for analysis in this chapter – I have visualized, organized, and adapted2 the ‘5 T’s of early literacy’ proposed by RTI (Bulat et al., 2017, p. 1). This scheme summarizes the many necessary conditions (and ongoing challenges) for early grade reading in developing countries. The adapted 5 T’s are Teaching, Texts, Tests, Training, and Tongue (language) (Figure 1). All of these factors are influenced by the context within which the school operates, particularly the cultural orientations and attitudes towards reading and education, as well as whether the minimum conditions for learning have been met (basic nutrition, non-extreme class sizes, essential learning materials, and so on). Before entering a discussion of these 5 T’s, it is worth emphasizing the primacy of the ‘minimum conditions for learning’ with a short sketch of a typical teacher’s situation. Illustrating the context within which many teachers must work helps to anchor expectations of what is possible given classroom realities in some countries. A 2014 nationally representative survey of 10 Francophone African countries, namely, Benin, Burkina Faso, Burundi, Cameroon, Congo, Côte d’Ivoire, Niger, Senegal, Chad, and Togo, found that 22% of primary school pupils were in schools that had no toilet facilities whatsoever (CONFEMEN, 2015, p. 104). In Togo, Chad, and Côte d’Ivoire, the figure is 40%. In Niger, the average Grade 2 class had 48 pupils in it, with much higher class sizes in Senegal (52 pupils per class), Benin (57 pupils per class), and a bewildering 80 Grade 2 pupils per class in Burkina Faso (CONFEMEN, 2015, p. 99). When the average Grade 2 Burkinabe teacher has 80 seven-year-olds in a single class, teaching more than a small minority of them to read is simply not possible. Assuming these elementary conditions are met, what are the factors that contribute to early literacy in developing countries? The first T in Figure 1 is Teaching, and its quality is determined to a large extent by the second T, Training. There is now a large body of consistent research revealing 2

INTRODUCTION

Figure 1. The 5 T’s of early grade reading (adapted from Bulat et al., 2017)

the essential components of teaching reading (Castles et al., 2018), supported by the latest findings in neuroscience (Dehaene, 2009; Seidenberg, 2017). These all foreground another ‘Big Five’ aspects of reading: (1) Phonemic Awareness, the ability to manipulate the smallest units of sounds (phonemes) in words; (2) Phonics, the systematic relationship between letters and sounds; (3) Fluency, the ability to read quickly and accurately; (4) Vocabulary, one’s breadth and depth of word knowledge; and (5) Comprehension, the ability to understand what is being read (Bulat et al., 2017; National Reading Panel, 2000). Unfortunately, these ‘scientific’ understandings of reading are rarely included in teacher training programs for either in-service or pre-service teachers. Both the length and the quality of initial teacher training will determine to a large extent whether teachers are equipped to teach reading. Turning again to the Francophone African context, two thirds of early primary teachers in Francophone Africa (64%), have received only one year, if that, of teacher training, with 22% having received no professional training whatsoever (CONFEMEN, 2015, p. 121). What further complicates this issue is that the vast majority of reading research has been conducted on English, which is by no means a typical language. This brings us to the third T, Tongue. As a field of researchers, we would do well to heed Pretorius’ admonition in the current volume, repeated in full below: 3

SPAULL

First, we need to be mindful of the sources of our knowledge about reading. From which languages do research findings derive? Second, given the complexity of reading, we need to be reasonably sure that there is converging evidence for the claims that are made about reading. Third, we need to be mindful of how we apply what we know and understand about reading in contexts that are different from the original contexts. Will the implied predictions or outcomes still hold true in a different context? Much of what we know and understand about reading comes from a large body of research based on reading in English and other European languages, among them, in particular, Dutch, Spanish, Italian, French, German, Greek, and Swedish. These are languages that belong to the Germanic or Romance language families. They are classified typologically on a continuum, according to whether they have greater or fewer analytic or inflectional morphosyntactic features (that is, whether they convey meaning primarily through syntax, especially word order, or by adding prefixes and suffixes to words). Increasingly, research into reading in languages that are typologically different, such as Turkish, Finnish, and Basque, is entering academic reading journals. These languages differ from European languages in that they are agglutinating languages. In these languages, words consist of sequences of morphemes attached to a stem, each of which represent a different grammatical category, as in the [isiXhosa] word abayifundi ‘they aren’t reading it’, which consists of five morphemes a+ba+yi+fund(stem)+i … Is reading in an agglutinating language the same as, or different from, reading in analytic or inflectional languages? (Pretorius, current volume) While Pretorius speaks from the South African context with African languages in mind, Sinha (current volume) shares similar but different concerns in relation to reading in Hindi (India): This vast body of research, as well as the building of theory, took place outside India in developed countries whose languages have alphabetic scripts. But Indian script, such as that of Hindi, is alphasyllabic. This raises questions. How can theories developed with a different type of script apply to Indian languages? (Padakannaya and Mohanty, 2004). (Sinha, current volume) In a similar vein, Menon and her co-authors (current volume) document in some detail the practical difficulties of teaching reading using Indian scripts, and specifically the question of how to sequence the acquisition of aksharas (letters of the Indian script) when teaching children to read. This issue speaks to the Text component of the 5 T’s – which texts, and what types of texts, should be used when teaching children to read? What their research shows is that the easiest words to write in Kannada (one of the Indian languages) are also words that are unfamiliar to children, while everyday words are often more difficult to write, since they include modifications to the aksharas in the form of maatras (secondary vowel diacritics). Thus, there is a trade-off between postponing word formation until children know 4

INTRODUCTION

how to use aksharas and maatras to form words they know orally, or quickly starting the forming of words whose meaning they do not know. Menon and her co-authors’ discussion is especially revealing: In contemporary Indian curricula, the emphasis is on forming meaningful words from the beginning. Hence, the order of introducing the base aksharas has been reorganized to permit the formation of words. However, curriculum designers hesitate to introduce symbols for maatras or conjunct consonant sounds too early – perhaps acknowledging the difficulties that young learners may have in simultaneously mastering so many different symbols of the script. Therefore, in order to make meaningful words, they need to find words that use only the inherent vocalic sounds in the base aksharas. Such words can, indeed, be made – but they tend to be words that are not commonly used by young children in their oral language, and are often words derived from the classical Sanskrit language – one of the languages of erudition and scholarship in earlier times. Therefore, the first words learned by a child in school tend to be those that are not in the spoken oral vocabulary of young children, and that represent objects and emotions that the child may have no organic relationship with. For example, the word ‘salaga’ (Sanskrit word meaning ‘tusker’) is easier to form from the aksharas introduced, than the word ‘aane’ (Kannada word meaning ‘elephant’ – which has a maatra). Likewise, the word ‘dhana’ (Sanskrit word for ‘wealth’) is easier to form than the word ‘rokka’ (dialectic word for ‘money’ in certain parts of Karnataka – which has a conjunct consonant symbol and a maatra in it). In our work, we observed a large number of specialized words being introduced to young children engaged in the task of akshara-mastery in their classrooms … While the surface features of the curricula have shifted – ostensibly towards meaning-making … the deep structures of teaching and learning literacy remain largely divorced from meaning-making. Thus, children aspiring to be first-generation literates in their families spend the better part of two years seeing pictures and copying down words representing tuskers and kings, chainsaws and inlaid necklaces – which have little or nothing to do with what they might actually want to learn or talk about. We asked a first grader why the words he used at home for different objects were different from the words he learned at school. He looked incredulously at us and responded: ‘At home, we speak in Kannada. At school, I am learning English!’ This first grader had not realized that the language he was learning at school was not, in fact, English, but school-based Kannada. (Menon et al., current volume) Issues raised by these and other authors in the current volume point to fundamental questions about the way literacy is conceptualised and taught in developing countries. Furthermore, these examples point only to the technical elements children battle as they try to become code-breakers, not to mention the challenges they face when becoming meaning-makers or text-critics (Luke and Freebody, 1999). Again, Menon and her co-authors show how cultural understandings of literacy can be especially 5

SPAULL

difficult to overcome. They show that in India, where texts are used in religious contexts for ritual purposes, meaning is ‘assumed to reside in the text and not in the minds of the readers. Individual meaning-making and interpretation were not significant to the functions or practices of this form of literacy.’ As a result, teachers tend to ‘favour shared interpretations of text over individual interpretations’. As the authors conclude, ‘When we speak of designing powerful reforms to literacy instruction, we’re seeking to accomplish nothing short of cultural change.’ It is these types of insights – and there are many in the current volume – that make clear the need for more research on early reading and writing in the languages children speak and the contexts within which they live in developing countries. Throughout the volume, the reader will see that authors engage with different components of the 5 T’s model. In the first chapter, McLachlan provides a helpful overview of the dominant ideas in early reading and concludes that “any literacy policy needs to be based on two things: on relevant language and literacy practices in African countries and on the evidence generated by research regarding the knowledge and skills that children need for literacy acquisition”. Chapters 2 and 3 deal with Tests and measurement from both a theoretical perspective (Afflerbach) and an applied perspective focussing on Burkina Faso and Senegal (Spaull and Lilenstein). Savage and Côté (in Chapter 5) look at Tongue and ask if there are relevant lessons to be learnt from the Canadian model of language immersion. The authors in Chapters 6, 7, 8, 9, and 10 discuss in some depth the Training element of early literacy, reflecting on experiences in India, Zambia, Chile and Argentina. The pioneering work of Pretorius and of Menon and her colleagues extend our understanding of how Tongue and Teaching work together to help or hinder early literacy. Finally, the authors of chapters 12 through 16 reflect on a multiyear intervention in Burkina Faso, Niger, and Senegal, foregrounding the importance of alignment across the 5 T’s. Given how little scholarly attention has been paid to this area of the world, these five chapters offer valuable insights into the practical challenges and opportunities of educational reform in developing countries. AN OVERVIEW OF EARLY GRADE READING OUTCOMES IN DEVELOPING COUNTRIES

Any discussion of early literacy in developing countries would be incomplete without a short overview of the nationally representative learning outcomes in these countries. Over the last 20 years, there has been a proliferation of cross-national assessments of reading outcomes (Gustafsson, 2018), allowing for international and inter-temporal comparisons of achievement. Figure 2 below shows the percentage of Grade 2 students who could not read a single word from a short text in the language in which they were being taught. In this selection of countries from Africa, Asia, and the Middle East, it is clear that at least a third of children, and in some countries (Malawi, India, and Ghana) as many as 80% of children are completely reading illiterate after two years of formal full-time schooling. 6

INTRODUCTION

Figure 2. Percentage of Grade 2 students who could not read a single word of a short text (Source: World Bank, 2018, p. 5, which uses nationally representative 2012–2017 data from the UWEZO and ASER tests and the Early Grade Reading Barometer, 2017).

Given this book’s special emphasis on Burkina Faso, Niger, and Senegal, it is worth further exploring the most recent cross-national assessment of reading in this region specifically. Periodically the Conference of Ministers of Education of French-Speaking Countries (CONFEMEN) conducts the PASEC3 assessment in a nationally representative sample of primary schools in Benin, Burkina Faso, Burundi, Cameroon, Congo, Côte d’Ivoire, Niger, Senegal, Chad, and Togo.4 The most recent assessment was in 2014. This study revealed that at the end of Grade 2, half (52%) of all the children assessed could not read 10 letters of the alphabet correctly in one minute. Only 25% could read more than 20 letters correctly in one minute (CONFEMEN, 2015, p. 38). In Benin, Chad, and Niger, more than two thirds of Grade 2 pupils do not know all of the letters of the alphabet (Figure 3). Looking again to the chapters in the current volume, Spaull and Lilenstein use household survey data together with results from literacy assessments to analyse a nationally representative cohort of Burkinabe and Senegalese children aged 11 to 15 years old. They find that only 23–34% of each cohort had both completed Grade 2 and acquired basic literacy (23% in Burkina Faso, and 34% in Senegal). For the poorest 40% of children the rates are 11–21%. Many of these results may be difficult to interpret for those in the West unfamiliar with the types of tests conducted in developing countries in Africa and Asia (such as the EGRA, ASER or UWEZO tests). In the few instances when developing countries do participate in the same international assessments with wealthier countries, the magnitude of the differences becomes truly apparent. In the 2016 round of the Progress in International Reading and Literacy Study (PIRLS), 78% of Grade 4 pupils in South Africa could not read 7

SPAULL

Figure 3. Percentage of Grade 2 students who do not know all of the letters of the alphabet (i.e., could not read at least 10 letters correctly in one minute) (PASEC, 2014). (Source: CONFEMEN, 2015, p. 38)

for meaning in any language (the PIRLS Low International Benchmark), compared to 69% in Egypt and 64% in Morocco (Mullis et al., 2017, p. 55). The international median across all 50 participating countries was 4%, with figures of 5% in Germany, 4% in the United States, and 3% in England. Given the hierarchical nature of reading and the fact that reading for meaning is a prerequisite for further engagement with the curriculum, it is to be expected that these stark differences at the Grade 4 level are even larger in higher grades. In 2015, a selection of developing countries participated in the well-known Program for International Student Assessment (PISA), which assesses 15-year-olds who are still enrolled in school. Fewer than 10% of 15-yearolds in school in Cambodia, Senegal, or Zambia were functionally literate (PISA Level 2), compared to the OECD average of 80% in 50 wealthy countries (OECD, 2018, p. 7). This is all the more concerning when one considers that in 2015 only a third of Cambodian, Senegalese, and Zambian 15-year-olds were actually still in school (OECD, 2018, p. 7). CONCLUSION: CHARTING THE PATH AHEAD

The education challenge facing the global community is immense. In addition to addressing the unfinished business of the Millennium Development Goals (equitable access and universal primary school completion), there are now the extra challenges of ensuring all children acquire ‘effective and relevant learning outcomes’, interpreted here as reading and writing for meaning and pleasure. The research findings documented in this volume help move the early literacy discussion forward in important ways by suggesting a number of new avenues for linguistic and pedagogical research. This includes the call for linguistically informed approaches to early literacy instruction; the alignment of training, materials, and assessment; and 8

INTRODUCTION

the need for the development of lexicons and grammars for under-studied languages. While this volume has contributed to our understanding of early literacy in these developing countries, it has also left us with more questions than answers. The only way to meet the education challenge posed by Sustainable Development Goal 4 will be to facilitate a shift in research priorities and funding towards developing countries: their languages, their contexts, and their policy challenges. There is good reason to do so. Ensuring that all children can read and write for meaning and pleasure will bear countless dividends for humanity as a whole. The benefits to education in general, and literacy specifically, are manifold and include lower fertility (Basu, 2002), improved child health (Currie, 2009), reduced societal violence and improved human rights (Salmi, 2000), increased economic growth (Hanushek and Woessman, 2008), promotion of a national – as opposed to a regional or ethnic – identity (Glewwe, 2002), and increased social cohesion (Heyneman, 2003). Over and above these singular benefits, learning to read for meaning and pleasure is arguably the best way to expand children’s capabilities and freedoms, enabling them to pursue the sorts of lives they have reason to value (Sen, 1999). NOTES 1

2

3 4

These were the Journal of Research in Reading, Scientific Studies of Reading, and Reading Research Quarterly. To identify these three journals, we use the SCImago Journal Rank Indicator, which calculates the average number of weighted citations received over the selected period. We limited the journals to those focusing specifically on reading and those that had an international focus. Note that these figures do not change substantially when one considers other leading reading journals and are only included here as illustrative of the broader challenge in international reading research. Across the three journals, the number of articles published on each language are as follows: English (730), French (240), Spanish (130), Portuguese (23), Chinese (210), Arabic (60), Hindi (9), Bengali (9), and Swahili (4). The original ‘5 T’s’ proposed by RTI were Teach, Text, Time, Test, and Tongue (Bulat et al., 2017, p. 1). In this model, ‘time’ is replaced with ‘training’, given how central teacher training is in almost all the recommendations of the authors in this collected volume, and Tongue is placed as an overarching connector across texts, tests, training, and teaching. CONFEMEN Programme for the Analysis of Education Systems. In the CONFEMEN-PASEC studies, tests are administered in the official language of instruction, which is French in all countries except in Anglophone Cameroon, where tests are administered in English, and in Burundi, where tests are administered in Kirundi (CONFEMEN, 2015, p. 20).

REFERENCES August, D., & Shanahan, T. (2006). Developing literacy in second-language learners: Report of the national literacy panel on language minority children and youth. Mahwah, NJ: Erlbaum. Basu, A. M. (2002). Why does education lead to lower fertility? A critical review of some of the possibilities. World Development, 30, 1779–1790. Bulat, J., Dubeck, M., Green, P., Harden, K., Henny, C., Mattos, M., … Sitabkhan, Y. (2017). What we have learned in the past decade: RTI’s approach to early grade literacy instruction. https://www.rti.org/sites/default/files/resources/14124740_RTI_Approach_Early_Grade_Literacy_ Instruction.pdf Castles, A., Rastle, K., & Nation, K. (2018). Ending the reading wars: Reading acquisition from novice to expert. Psychological Science in the Public Interest, 19(1), 5–51.

9

SPAULL CONFEMEN (2015). PASEC 2014, education system performance in Francophone sub-Saharan Africa: Competencies and learning factors in primary education. Dakar: CONFEMEN. https://www.pasec.confemen.org/wp-content/uploads/2015/12/Rapport_Pasec2014_GB_webv2.pdf Currie, J. (2009). Healthy, wealthy, and wise: Socioeconomic status, poor health in childhood, and human capital development. Journal of Economic Literature, 47(1), 87–122. Dehaene, S. (2009). Reading in the brain: The new science of how we read. London: Penguin. Ethnologue (2019). Ethnologue languages of the world. https://www.ethnologue.com/browse/names Glewwe, P. (2002). Schools and skills in developing countries: Education policies and socioeconomic outcomes. Journal of Economic Literature, 40, 436–482. Gustafsson, M. (2018). Costs and benefits of different approaches to measuring the learning proficiency of students (SDG Indicator 4.1.1). Paris: UNESCO. Hanushek, E. A., & Woessmann, L. (2008). The role of cognitive skills in economic development. Journal of Economic Literature, 46(3), 607–668. Heyneman, S. P. (2003). Education, social cohesion, and the future role of international organizations. Peabody Journal of Education, 78, 25–38. Luke, A., & Freebody, P. (1999). Further notes on the four resources model. http:www.readingonline.org/ research/lukefrebody.html Mullis, I. V. S., Martin, M. O., Foy, P., & Hooper, M. (2017). PIRLS 2016: International results in reading. Boston, MA: International Association for the Evaluation of Educational Achievement. Muralidharan, K. (2013). Priorities for primary education policy in India’s 12th five year plan. India Policy Forum 2012–2013, 9, 1–46. National Reading Panel (2000). Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction. Washington, DC: US Government Printing Office. OECD (2018). PISA for development: Results in focus. PISA in Focus, 91. https://www.oecd-ilibrary.org/ education/pisa-for-development_c094b186-en Padakannaya, P., & Mohanty, A. K. (2004). Indian orthography and teaching how to read: A psycholinguistic framework. Psychological Studies, 49, 262–271. Piper, B., Zuilkowski, S., Kwayumba, D., & Oyanga, A. (2018). Examining the secondary effects of mother-tongue literacy instruction in Kenya: Impacts on student learning in English, Kiswahili, and mathematics. International Journal of Educational Development, 59, 110–127. Piper, B., Zuilkowski, S., & Ong’elle, S. (2016). Implementing mother tongue instruction in the real world: Results from a medium-scale randomized controlled trial in Kenya. Comparative Education Review, 60(4). Pritchett, L. (2013). The rebirth of education: Schooling ain’t learning. Washington, DC: Center for Global Development. Salmi, J. (2000). Violence, democracy and education: An analytical framework. Washington, DC: World Bank. Seidenberg, M. (2017). Language at the speed of sight: How we read, why so many can’t, and what can be done about it. New York, NY: Basic Books. Sen, A. (1999). Development as freedom. Oxford: Oxford University Press. Snow, C. E. (2017). Early literacy development and instruction: An overview. In N. Kucirkova, C. E. Snow, V. Grøver, & C. McBride-Chang (Eds.), The Routledge international handbook of early literacy education: A contemporary guide to literacy teaching and interventions in a global context (pp. 5–13). Abingdon & New York, NY: Routledge. Spaull, N., & Taylor, S. (2015). Access to what? Creating a composite measure of educational quantity and educational quality for 11 African countries. Comparative Education Review, 59, 133–165. Taylor, S., & von Fintel, M. (2016). Estimating the impact of language of instruction in South African primary schools: A fixed effects approach. Economics of Education Review, 50, 75–89. http://dx.doi.org/10.1016/j.econedurev.2016.01.003 UNESCO (2017). Unpacking Sustainable Development Goal 4, Education 2030. Paris: UNESCO. World Bank (2018). World development report 2018: Learning to realize education’s promise. Washington, DC: World Bank. doi:10.1596/978-1-4648-1096-1

10

CLAIRE MCLACHLAN

1. THE EARLY READING CURRICULUM International Policy and Practice

INTRODUCTION

Using research conducted in well-developed Western nations to propose strategies for less-developed, socially-and-linguistically-dissimilar countries is inherently dangerous. Research from developing countries, however, is comparatively scarce. Indeed, according to Pence and Ashton (2016), although 15% of the world’s children live in sub-Saharan Africa, only a negligible amount of published early years literature focuses on the region. These authors point out also that research within sub-Saharan Africa is currently on the rise, and that this in itself will make Africa increasingly able to define its own way forward. This chapter, while bearing in mind the foregoing caveat, nevertheless draws on policies and practices that have proved useful in Western countries, to put forward a range of factors potentially relevant to devising approaches to early literacy in developing countries. It will examine the need for literacy policy and explore Olson’s (2009) notions of ‘Read what?’ and ‘How well?’ – notions that underpin many literacy policies. It will then look into the importance of having policies on reading-writing acquisition and differentiated instruction in the junior primary classroom, and will examine, from different theoretical perspectives, the conditions essential to promoting early years literacy. The chapter will then move on to explore the issues related to transitioning from the early childhood context to the primary context, with a specific focus on supporting children’s learning outcomes. Its final section focuses on the issues associated with studying early literacy education from a systemic perspective, and examines corresponding implications for local and national policy development. POLICY FOR EARLY READING: DOMINANT IDEAS

It can be argued that most countries want their children to achieve literacy and to be able to participate in global society. This section will explore some dominant ideas around literacy policy in recent years, examining how these shape literacy policy for beginning reading in different countries. Reading and writing skills are key to achievement in most curricula and are typically the foundation for learning in all subjects studied at school. If children do not learn to read, their general knowledge, spelling, writing, and vocabulary development suffer (Stanovich, 2000).

© JOINTLY BY IBE-UNESCO AND KONINKLIJKE BRILL NV, LEIDEN, 2019 DOI: 10.1163/9789004402379_002

MCLACHLAN

However, in many countries, ideology plays a critical role in the development of literacy education policies (Allington, 2010). As a result, decisions about how to achieve literacy in children can differ widely from country to country. Moreover, as Triebel (2005) points out, countries’ perceptions of what counts as literacy will also differ considerably, often depending on their level of economic development. This section will explain the recent history of literacy policy in developed and developing countries, and will also explore the ideologies underpinning current approaches to literacy in many countries. Triebel (2005) argues that in both developed and developing countries there have been three main drivers of literacy policy in the last half century. The first driver is the politics of education, according to which the democracy and development of a nation is delivered via the written word and illiteracy is an injustice (Freire and Macedo, 1987). The second is an economic rationale of literacy, which dictates that investment in education will lead to societal improvements and economic and political stability, and which therefore heavily influences international development agencies such as the World Bank and OECD. The third driver is the strategy for schooling advocated by UNESCO in its campaigns for the eradication of illiteracy and adopted by the World Bank and UNICEF. The resulting literacy policies reflect the view that education is the path to personal, social, and economic development, and they focus on alphabetical skills as integral to education. However, as Triebel also points out, in the 1990s a change in focus led to recognition of the fact that these large-scale approaches to literacy in developing countries were failing. It became apparent that to achieve literacy, literacy processes needed to be socially and institutionally embedded in a region’s culture and collective identity; that achieving literacy was not only about implementing approaches to build literacy competencies, but also about creating the structural conditions necessary for literacy to become an imperative to a community and an economy. Recent research into literacy draws on these two strands and falls into two dominant ideologies, which in turn drive current literacy policy. These ideologies are based on perspectives of literacy derived from either social practice (e.g., New London Group, 1996; Kalantzis and Cope, 2012) or cognitive science or neurolinguistics (e.g., Tunmer, Chapman, and Prochnow, 2006; NELP, 2009). Allington (2010) states that, to a large extent in both developed and developing countries, policy to improve national literacy outcomes has been driven more by ideology than by research, and that it draws on the different social practices surrounding literacy development in different countries. As I have argued elsewhere, to serve as a coherent basis upon which to develop literacy policy, an approach should integrate both social practice and cognitive/neurolinguistic research (McLachlan and Arrow, 2011; McLachlan, Nicholson, Fielding-Barnsley, Mercer, and Ohi, 2013), but the significance of the difference needs to be explored. The social practice perspectives propose that literacy should be conceived of as multiliteracies, which are shaped by social diversity or by the variability of conventions of meaning in different cultural, social, and domain-specific situations, 12

EARLY READING CURRICULUM

and by multimodality, in which written-linguistic modes of meaning interface with oral, visual, gestural, and spatial patterns of meaning (Kalantzis and Cope, 2012). Although this view of literacy does not negate the importance of learning spelling and grammar rules, it suggests that being a successfully literate person in a globalized world requires additional functional and communicative skills. Policies that promote literacy as a social practice therefore operationalize literacy by gearing literacy education towards engagement and social interaction (Olson, 2009). In this view, what counts as literacy and what is helpful or useful literacy practice depend on the situation or context. This social practice view of literacy is reflected in the UNESCO report prepared by Awopegba, Oduolowu, and Nsamenang (2013), which advocates the development of early years curricula based on local approaches to language and literacy. The cognitive science perspective is based on the body of research that examines the relationship between ‘literate cultural capital’ (Tunmer and Chapman, 2015) on school entry (‘literate cultural capital’ is defined as the literacy knowledge, skills, and experiences children have acquired in their homes and early childhood settings) and the effectiveness of different approaches to teaching literacy in terms of children’s educational outcomes and life chances. The research in this field is typically evidence-based and examines the interface between educational policies and scientific research on reading. The cognitive science/neurolinguistics view draws on research that has examined which factors predominate in children’s literacy acquisition. An example of how this perspective has been operationalized is the ‘No Child Left Behind’ policy of ‘Reading First’, which is based solely on scientific, reliable, replicable reading research (SBRR) (Allington, 2010). Differences in the levels of knowledge and awareness that children have during their preschool years can affect the efficiency with which they transition into conventional literacy in formal settings (Tunmer, Chapman, and Prochnow, 2006). Experiences in early childhood are therefore of paramount importance. Research shows that children’s experiences of literacy in their homes and communities powerfully influence how easily they develop literacy in educational settings (Sénéchal, 2011). The work of Sénéchal, for example, in showing that by engaging in advanced literacy play (beyond simply story reading) parents can enhance their children’s literacy development, demonstrates the significance of both cognitive and social approaches to literacy acquisition. There are, of course, multiple literacies that children experience in their homes, communities, and cultures, all of which shape the ways in which they experience literacy (Makin, Jones Díaz, and McLachlan, 2007; McLachlan et al., 2013). Literacies are experienced both inside and outside formal settings (Knobel and Lankshear, 2003) and within culturally specific contexts (Tagoilelagi-Leota, McNaughton, MacDonald, and Ferry, 2005). Some of these contexts are aligned with conventional school-based literacies, while others are not. A coherent literacy policy should enable differentiated instruction so as to ensure that all children, on entering school, are afforded opportunities to build on the literacy knowledge and skills they developed in early childhood. 13

MCLACHLAN

Olson (2009), who has written extensively on the links between educational theory, literacy policy, and literacy acquisition, states that the primary definition of a literate person is one who is able to read and write, and that literacy is a basic, personal competence. He states that while basic literacy, that is, the ability to read and write, is mastered fairly readily by most children, the ability to use literacy more powerfully for different purposes, within differing social practices and according to different conventions, requires higher levels of understanding and mastery. In order to turn this insight into policy, Olson says that governments need to deal with two specific questions: (1) How well? (which examines standards); and (2) Read what? (which examines content). Olson (2009) argues, then, that governments have two policy development options: to develop policies with very narrow, precise goals, thus facilitating the measurement of all levels of achievement; or to develop policies with broader goals, a possible consequence of which is that not all children will achieve them. Curriculum documents and literacy policy around the world differ in terms of expectations of literacy achievement. This state of affairs undoubtedly shapes teaching practices in junior classrooms. An example of a strategy based on the question ‘How well?’ is the British Foundation Stage curriculum (Department of Education, 2014), which assesses children entering school against 17 specific criteria. This allows for a very precise measurement of their literacy achievements. The British Foundation Stage (Department of Education, 2014) curriculum specifies that children will learn in three prime areas: communication and language; physical development; and personal, social, and emotional development. Children are also expected to learn in four specific areas of the curriculum: literacy, mathematics, understanding of the world, and the expressive arts and design. Literacy is defined as follows: ‘Literacy development involves encouraging children to link sounds and letters and to begin to read and write. Children must be given access to a wide range of reading materials (books, poems, and other written materials) to ignite their interest’ (p. 8). However, the curriculum goes beyond simply inciting interest and encouraging children to play with letters and sounds and explore writing, which is implicit in other countries’ policy documents. Each child has a profile prepared for them by the time they are five years old. This profile documents the teachers’ assessments of whether the child is meeting, exceeding, or not yet reaching, the expected levels of development. The purpose of assessing the children against the aforementioned criteria is to establish where they need further support physically, cognitively or socio-emotionally. Similarly, Li and Rao (2005) discuss the differences in policy for early childhood literacy in Beijing, Hong Kong, and Singapore. In Beijing, teaching literacy is banned until children are six. The policy in Hong Kong is to promote a child-centred curriculum. But this approach is often ignored by parents there, who want their children to be intensively drilled in the English language. Singapore has a compulsory bilingual policy, but English is taught as the first language. In all cases, there are clear expectations of what children should be able to do at specific ages, although 14

EARLY READING CURRICULUM

there may be differences in school and family expectations. Allington (2010), reflecting on ‘How well?’ strategies, argues that similar approaches in the United States (implemented in response to the federal push to incorporate high standards, test-based accountability and specified models of curriculum and instruction) have failed to close achievement gaps because the corresponding policies are based on ideology rather than wide evidence. The New Zealand education system provides us with an example of a more broadly framed ‘Read what?’ approach. While children typically start school on their fifth birthday, their literacy achievement is not assessed until they are six (the official school entry age). This suggests a ‘wait to fail’ approach to literacy instruction (Tunmer and Chapman, 2015), which is equally problematic. With this approach to reading, children potentially cultivate poor reading habits for a whole year at school before any intervention is offered, making it more likely that they will fail to learn to read. The national curriculum in New Zealand (Ministry of Education, 2007) is quite explicit about how children will develop literacy at school. It highlights that children need to make connections between letters and sounds, and that they will slowly develop a sight-word reading vocabulary and knowledge of text conventions within a rich language and literacy learning environment. The New Zealand’s Literacy Learning Progressions (Ministry of Education, 2012) provide further information for teachers, specifying the levels of knowledge and skill children are expected to have at school entry, and defining the milestones they should have reached by particular times (for example, after one year of study). The progressions are based on a model that stipulates that there are three aspects to literacy acquisition: ‘learning of the code’, making meaning, and thinking critically. Assessing progress against ‘learning of the code’ involves examining children’s development in terms of their having ‘an awareness of rhyme and words that start with the same sound’, ‘the ability to read their own name’, ‘some concepts about print’, and ‘an ability to identify the first letter of their name and some other letters’ (Ministry of Education, 2012, p. 9). Children are also expected to be familiar with, and show engagement in, story reading. However, 98% of all children attend an early childhood service in New Zealand, and therefore teachers can anticipate that many children will have had these experiences and understandings before school entry. Olson also argues that children need to gain an understanding of societal literacy in order to function fully and effectively and that most literacy policymakers should focus their attention on the issues of ‘Read what?’, ‘How well?’ and ‘For what social purpose?’. Olsen (2009) proposes that many countries are promoting ‘societal literacy’ as part of determining what skills children need for social participation, and that these aspirations underpin government policy. Although most countries espouse children’s rights to education and literacy acquisition at a national level, there also needs to be commitment at a local level to supporting bilingualism and biliteracy. Literacy policy needs to draw on the growing body of research on supporting bilingualism and biliteracy (Tagoilelagi-Leota et al., 2005; Tabors and Snow, 2001), and the need for culturally and linguistically responsive classrooms (Gay, 2010). 15

MCLACHLAN

Countries that favour either the social practice or the cognitive science approach to literacy, rather than a more balanced, inclusive approach, face certain problems inherent in their choices. As Spencer, Falchi, and Ghiso (2011) argue, early literacy policy cannot be deterministic, because outcomes of educational approaches cannot be precisely determined, and the early literacy curriculum cannot be standardized, because of the diverse needs of learners. Instead it needs to be a negotiation between school and community that supports multilingual learners. As argued elsewhere (McLachlan et al., 2013), children are active drivers of their own literacy acquisition, and their learning journey is shaped by biology and by social, cultural, and educational factors. Similarly, Stahl and Yaden (2004, p. 141) argue that teachers and researchers need to: … probe more deeply into research emanating from biology and the neurosciences (see Shonkoff and Phillips, 2000), children learning English as a second language (August and Hakuta, 1997), learning theory (Bransford, Brown, and Cocking, 1999), and investigations into the causes of reading difficulty (Snow, Burns, and Griffin, 1998). Taken together, these findings point to the fact that all cognitive and affective learning processes are highly complex interactions between inherited and environmental factors and are selectively affected by variations in child-rearing practices, socioeconomic circumstances, family structures, adult-child interactions, educational environments, and other contextual and developmental factors. As this brief discussion indicates, many countries take different approaches to answering the questions ‘Read what?’ and ‘How well?’ (Olson, 2009). These approaches range from completely banning literacy teaching for five-year-olds to rigid literacy standards for five-year-olds. However, assessing literacy acquisition at five or on school entry is a difficult undertaking. After all, literacy develops on a continuum (Whitehurst and Lonigan, 1998). Moreover, young children are notoriously challenging to accurately assess, with their high distractibility, high energy levels, possible concentration issues, low frustration tolerance, and little social interest in pleasing an assessor, parent or teacher (McLachlan, Edwards, Margrain, and McLean, 2013). How to achieve literacy outcomes for children is a complex question. What is clear, however, is that policy needs to be relevant to the children and to take into account their previous literacy experiences. The next section will examine what the essential dimensions of an early reading curriculum should be to answer Olsen’s (2009) question ‘Read what?’ It will also examine issues related to his question ‘How well?’ PROMOTING LITERACY IN THE JUNIOR PRIMARY CONTEXT

This section addresses some of the essential elements that an early reading curriculum should take into account. These include the need for a literacy-rich environment and the importance of focusing on teaching fundamental literacy knowledge and 16

EARLY READING CURRICULUM

skills, particularly in developing countries where exposure to literacy experiences during childhood may be limited. However, before the essential dimensions of the early reading curriculum can be examined, the early childhood context from which children will transition to school needs to be considered. Whitehurst and Lonigan’s (1998) seminal definition of emergent literacy is useful when conceptualizing what teachers need to know about literacy for children on school entry and to answer the questions ‘Read what?’ and ‘How well?’. They use ‘emergent literacy’ to ‘denote the idea that the acquisition of literacy is conceptualized as a developmental continuum, with its origins early in the life of a child, rather than an all-or-none phenomenon that begins when children start school’ (p. 848). Emergent literacy means children develop reading, writing, and oral language concurrently and interdependently as a result of exposure to social contexts and without formal instruction (Whitehurst and Lonigan, 1998). A child’s language development and family literacy patterns are also strong predictors of literacy achievement (Sénéchal, 2011). Where children have not had strong exposure to a wide range of literacy experiences, they will need such opportunities on school entry to provide a foundation for literacy instruction. Children entering school with minimal literacy-related experiences and/or low levels of literacy-related knowledge and skills will, of course, also require their early reading curriculum to include a focus on developing such knowledge and skills in order to take full advantage of subsequent reading and writing instruction within a class setting. Children who are most at risk of not learning to read are those who have not developed the school-type behaviours and understandings of how print works in either the home or an early childhood setting (Heath, 1983; Sénéchal, 2011) and will need a curriculum that bridges the differences in how literacy is practiced between home and school. Children who have not had the benefit of early childhood education (ECE) or have had limited ECE are likely to need targeted emergent literacy experiences in order to gain the foundational knowledge and skills needed for literacy acquisition. These children will need a literacy-rich early curriculum that focuses on the development of emergent literacy knowledge and skills. This is, however, easier said than done in schools facing multifactor challenges, such as those present in many African countries. Imoh and Ansell (2013) present a compelling case for the United Nations Convention on the Rights of the Child (UNCRC) and the importance of the African Charter on the Rights and Welfare of the Child (ACRWC). They argue that this document serves as a lens through which to gain global perspectives on children’s rights in Africa, which in turn have implications for literacy policy and practice. Awopegba, Oduolowu, and Nsamenang (2013) note that although 26 of the African countries have included early childhood education in their sector or national development plans, in most countries fewer than 20% of children actually attend an early childhood service. In many countries, the figure is well below 10%. According to a UNESCO BREDA (2012) estimate, more than one in two children receive no early childhood education. These figures underscore the importance of providing 17

MCLACHLAN

specific guidelines for supporting children in the junior primary classroom if children are to become literate. Cleghorn and Prochner (2010), citing the studies of LeVine et al. (1994) into childrearing among the Gusii people of Kenya and among white middle-class families in the United States, suggest that there are two dominant modes of ECE: the pediatric model and the pedagogical model. The two models are discussed in relation to the value a society attaches to interpersonal relatedness versus independence; emotional interdependence versus autonomy; material interdependence versus material emancipation; and collective versus individualistic autonomy (Kagitcibasi, 2007). Cleghorn and Prochner propose that the models of socialization implicit in these approaches persist into later teaching and learning situations, and that each model is adapted to a group’s need for survival. More specifically, they argue, when the economy is based on agricultural subsistence and the home environment is relatively non-literate, the model of childcare is more likely to be pediatric than pedagogical. More explicitly, the models of socialization that children experience as part of pediatric childcare are likely to be based on oral storytelling and peer interaction rather than adult-child interaction, and on the mastery of skills through imitation and observation. Under such circumstances, children will find a primary school environment unfamiliar and will have many cultural, linguistic and conceptual barriers to contend with. Although the pediatric model of child rearing is beneficial in terms of ensuring the survival of communities, it means that children are poorly equipped for the challenges of a primary school classroom and that they may have poor literate cultural capital (Tunmer and Chapman, 2015) with which to enter a beginning reading programme. Children who enter primary school with high levels of literate cultural capital (including knowledge of book grammar, print concepts, and basic phonological awareness) have often been exposed to many literacy related experiences in their homes and early childhood settings, such as listening to stories, playing rhyme games, learning nursery rhymes and jingles, playing word games, learning and singing the alphabet song, and writing and/or reading some simple sight words (Evans and Shaw, 2008; Phillips and Lonigan, 2009; Sénéchal, 2011). Such experiences are common in the pedagogical model of childcare (Cleghorn and Prochner, 2010). Children who have had this explicit and implicit exposure to literacy have the foundational emergent literacy skills that enable them to begin learning to read efficiently (Whitehurst and Lonigan, 2001), regardless of the approach to reading (Connor, Morrison, and Katch, 2004). Considerable research has already been conducted into how to support early years education in Africa (Awopegba et al., 2013). It has examined, for example, what sort of early childhood curriculum children should experience and how this should relate to African cultural norms and to the interactions between the developing child and their social/cultural context – something Vygotsky (Vygotsky, Rieber, and Carton, 1987) terms the ‘social situation of development’. Pence and Ashton (2016) refer to the extensive findings arising from the work of Garcia and colleagues (Garcia, 18

EARLY READING CURRICULUM

Pence, and Evans, 2008) and the edited volume Africa’s future, Africa’s challenge: Early childhood care and development in sub-Saharan Africa (2008). Although directions for policy were highlighted in this volume and the subsequent work, Pence and Ashton (2016) emphasize the challenges of building teachers’ capacity to implement policy and support children’s learning. Ifakachukwu (2011, p. 30) identifies ‘a yawning gap between policy formulation and implementation’. Pence and Ashton (2016) note the clear impact of infrastructural issues on policy implementation. These include the availability of adequate facilities and of pedagogical aids such as learning materials, and issues related to the salary and working conditions of teachers. Proposed solutions in Africa and other countries facing the same issues have included standardizing the curriculum to guide initial teacher education, focusing resources on the most disadvantaged children and implementing flexible models of field-based teacher education. Awopegba, Oduolowu, and Nsamenang (2013) note some principles that need to be recognized in the development of policy for education in the early years, which are of particular relevance to the teaching of reading and writing. These include Africa-focused pedagogies and local learning materials, a ‘generative curriculum’ (one arrived at with expert and community input) and locally influenced models of teaching and learning, community ownership and adequate resourcing and material support. These principles provide a useful social practice foundation for promoting literacy in the junior primary classroom. How to teach literacy to children on a wide developmental continuum will be explored next. ACHIEVING LITERACY OUTCOMES IN THE EARLY READING CURRICULUM

Walpole, Justice, and Invernizzi (2004) specify five areas that research has identified as key to literacy success for children. These include teachers knowing the predictors of reading achievement and identifying vulnerable children; understanding individual differences; knowing different approaches to instruction and how to prepare effective learning environments; and understanding school and curricular variables that influence achievement. This section examines some of the conditions an early literacy curriculum needs in order to foster effective learning and achieve learning outcomes. This section includes an analysis of what the teacher must know and understand, along with some examination of literacy policy for early reading. Also important is to look at what is needed in an effective early literacy curriculum. Although the research evidence is primarily Western, some key findings can potentially be applicable in developing countries. Providing children with varied literacy opportunities (for example, rich vocabulary, knowledge of the alphabet, awareness of sounds in the language, and an understanding of the purposes, functions, and structures of print) in the junior primary classroom provides a robust foundation for language and literacy development. A considerable body of research examines the importance of a rich literacy environment (Teale et al., 2009). However, children also need opportunities 19

MCLACHLAN

to interact with adults or with other more competent children to learn how to use the knowledge they acquire and make the most of the opportunities provided. As children may not have a solid grasp of the alphabetic principle on school entry (owing to adults having lacked a literacy focus in their child-rearing approaches or to the children’s limited exposure to ECE), it is important that time be spent with children on a range of emergent literacy tasks such as story reading, singing songs and nursery rhymes, and using language for a range of purposes. Teachers will need to provide children with a range of emergent literacy experiences. They will also need to understand and promote second language and literacy acquisition. Arguably, formal literacy instruction should be built on an enriched early literacy environment as children enter school. However, a rich literacy environment is not solely dependent on physical resources: the key resource is the teacher and the teacher’s ability to build a language- and literacy-rich environment (Morrow, 2009). Drawing on the evidence from international reading research, the National Early Literacy Panel report (NELP, 2009) identified some critical literacy understandings that children need in order to become literate at school. These include knowledge of the alphabet, phonological awareness (being aware of sounds in words), the ability to rapidly name letters, numbers, objects, and colours, the ability to write their own name, and the ability to remember spoken information for a short period of time. In addition, children need to understand print conventions and concepts, have strong oral language, and be able to match and discriminate visual symbols (NELP, 2009). Of these critical understandings, knowledge of the alphabet and phonological awareness are especially important and should therefore be fundamental to the early reading curriculum. Understanding letters and sounds forms part of the ‘inside-out’ processes that comprise emergent literacy (Whitehurst and Lonigan, 1998). But, while necessary, they are not sufficient for the acquisition of literacy (Muter, 1994). The ‘outside-in’ domain is also important. It represents children’s understanding of information outside of the particular printed words they are trying to read. It depends on knowing meanings of words, conceptual knowledge of the subject, and concepts about print. The inside-out units, on the other hand, represent children’s knowledge of the rules for translating print into spoken words. Each ‘domain’ has a different role to play in literacy, but together they form the basis for the acquisition of the ‘alphabetic principle’, which is the understanding that speech sounds in spoken words are represented by graphemes in print (Moats, 2000). This combined knowledge enables children to use letters and their sounds to make phonemically correct representations of words when both reading and spelling (Nicholson, 2005). Children with greater phonological awareness at school entry tend to be better readers (Stanovich, 2000); opportunities to gain phonological awareness in the junior primary classroom are therefore imperative. In addition, according to Ehri (2005), even very young children are able to learn to read new words. Once they acquire letter knowledge and some phonemic awareness, they move from a pre-alphabetic phase to a partial alphabetic phase, and new words are stored in memory, making use of partial connections between letters and sounds 20

EARLY READING CURRICULUM

for storage and for attempting unknown words with those connections. Boyer and Ehri (2011) state that although spoken language is acquired easily, children require explicit teaching to learn how to segment and blend phonemes in words, because speech is perceived as an unbroken flow of sound with no pauses between phonemes in words. This suggests that providing opportunities to gain alphabet and word knowledge as well as phonological awareness should be an important area of focus in the junior classroom. Intervention studies that have involved teaching phonological awareness skills to children prior to school entry show that such skills create advantages for children (Phillips, Clancy-Menchetti, and Lonigan, 2008). Recent studies show that literacy knowledge and skills can be promoted within naturalistic settings using a literacyrich environment and ‘teachable moments’ (Bruner, 1960) that comprise a range of effective literacy pedagogies and one-on-one instruction, and that last no more than 15 minutes per episode (Phillips et al., 2008). It can be argued that teachers’ knowledge about literacy in junior primary classrooms, especially for children who have had limited ECE literacy experiences, is of paramount importance (McLachlan and Arrow, 2015). Teachers will need explicit understandings of how language and literacy develop, of what influences different family and cultural contexts exert, and of how they will need to be able to use a range of pedagogies for supporting literacy learning. In primary classrooms in which a single teacher is responsible for the delivery of the curriculum, effective literacy practice will be largely based on the teacher’s knowledge because, in these circumstances, most instructional decision-making is an individual responsibility (Allington, 2010). Cunningham, Perry, Stanovich, and Stanovich (2004) state that literacy domain knowledge is crucial for teachers of young children. They argue that ‘there are strong theoretical reasons to suspect linkages between teacher knowledge and the ability to teach reading effectively (for example, being able to teach phonemic awareness and to choose good literature)’ (p. 160). It is imperative that teachers have a strong understanding of diverse learning pathways and of the many ways in which bilingual and multilingual children acquire and display literacy knowledge and skills. Relevant teacher knowledge includes knowledge of language and linguistics, an understanding of the psychology of reading development, and knowledge of effective practices that are based on research-informed models of cognitive development (Moats, 1994; Arrow, McLachlan, and Greany, 2015). Research suggests, however, that many teachers feel unprepared for teaching beginning reading (Buckingham, 2013), and in addition, Allington (2010) reports that only 25% of teachers are effective teachers of children who are struggling to learn to read. Wong Fillmore and Snow (2000) argue that teachers need intensive preparation in ‘educational linguistics’, because they gain inadequate preparation in most teacher education programmes. The case for early intervention to prevent long-term reading difficulties has been well documented (Tunmer and Chapman, 2015), but effective intervention is based on explicit teacher knowledge of research-based literacy assessment methods and effective pedagogies 21

MCLACHLAN

for differentiated instruction (Piasta, Connor, Fishman, and Morrison, 2009; Spear-Swerling and Zibulsky, 2014), and on teachers’ ability to apply this knowledge to assessment of children’s level of progress on the literacy continuum. Moats (2009, p. 393) points out that teacher knowledge is crucial for the implementation of differentiated instruction, as such instruction ‘depends on the teacher’s ability to explain concepts explicitly, to choose examples wisely, and to give targeted feedback when errors occur’. Internationally, some research exists to suggest that professional development can change teachers’ practices and children’s literacy outcomes, but these outcomes are influenced by the time and intensity of the intervention (Wasik, Bond, and Hindman, 2006). Ideally, teachers learn how to teach literacy in their initial teacher education, rather than having to rely on professional learning once they are in the profession. A SYSTEMIC APPROACH TO LITERACY IN THE EARLY YEARS

The notion of systemic approaches to learning stem from Ausubel’s (Ausubel, Novak, and Hanesian, 1978) theory of meaningful verbal learning, in which experiences are presented schematically to demonstrate connectivity, interactivity, and interdependence between old and new experiences (Al-hajaya, 2012). The theory suggests that learning results from the interaction of new knowledge with old knowledge – what Vygotsky (Vygotsky et al., 1987) called ‘schooled’ or ‘scientific’ knowledge – in ways that enable learning to develop as an organized, coherent, and logical structure. This notion is consistent with notions of building on children’s ‘funds of knowledge’ (Moll, Amanti, Neff, and Gonzalez, 1992) and with the approaches to learning advocated by Awopegba, Oduolowu, and Nsamenang (2013) for learning in the early years in Africa. Systemic approaches also draw on constructivist theorizing, in which learning is seen to occur as new concepts are integrated with existing understandings. McLachlan et al. (2013, p. 175) advocate that there needs to be a coherent framework for the construction of a junior school literacy programme. The five components of this framework are as follows: ‡ Building on children’s ‘funds of knowledge’. Drawing on the work of Moll et al. (1992) and on Thomson’s (2002) notion of a ‘virtual school bag’, this strategy involves finding out what literacy knowledge and practices children bring with them on school entry. Teachers may use activities to identify what key understandings children have at that time. Such understandings include recognition of environmental print, phonemic awareness, book orientation, ability to retell a story, awareness of certain concepts about print, and ability to write one’s own name. ‡ Building partnerships (teachers, child, and family). Teachers need to identify and acknowledge family literacy practices and work in partnership to support children’s literacy on school entry by discussing a range of effective strategies 22

EARLY READING CURRICULUM

for use at home and at school. The work on family literacy is particularly useful in this regard. ‡ Maintaining ongoing records of the child’s literacy development. Teachers need to look for signs of key accomplishments being achieved, such as understanding the alphabetic principle, reading sight words, reading words by mapping speech sounds to parts of words, and achieving reading fluency and comprehension. Teachers need to use a range of assessment methods, including event-based observations, running records, recordings of children speaking, and artifacts of children’s work, such as writing. Ongoing assessment is needed to plan for differentiated instruction (Arrow, 2015). ‡ Taking into account key understandings about children’s literacy learning in the early years. It is important that teachers understand that children do not all learn at the same pace and that there are continuities and discontinuities in children’s acquisition of literacy, many of which stem from ongoing differences in children’s home environments and social practices. Teachers need to understand that they are helping children to move from home concepts to ‘schooled’ concepts of literacy learning (Vygotsky, 1987). ‡ Providing engaging strategies and activities that support children’s learning. McLachlan et al. (2013, p. 183) advocate a range of activities and teaching strategies for supporting children’s acquisition of emergent literacy knowledge and skills which involve alphabet knowledge, phonological awareness, vocabulary development, building comprehension, writing, and handwriting. Morrow (2009) has written about the purposes, organization, and outcomes of literacy environments for young children, and she has specific guidance about what needs to be included in the classroom. She argues that children who have access to a literacy centre read and look at books 50% more often than children who do not have a specific space for literacy. Within the material support needs to be the provision of resources for each classroom, including a range of reading materials that children can use as they learn to read and write – materials at appropriate developmental levels and in appropriate languages and which depict images of the local community. Materials need to also include the range of resources that are typically available in a language- and literacy-rich learning environment. Note that in this increasingly digital age and to aid the development of multiliteracies knowledge and skills, it is also important to consider the use of digital resources such as iPads with e-books and e-games loaded for children’s use. Although these digital resources are not fundamental to learning to read and write, in an increasingly globalized and digitalized world, it is an increasingly prevalent aspect of literate cultural capital (Tunmer and Chapman, 2015). As Awopegba, Oduolowu, and Nsamenang (2013) argued, however, any approach also needs to include Africa focused pedagogies, local learning materials, a ‘generative’ curriculum (one arrived at with expert and community input), locally influenced models of teaching and learning, community ownership, and adequate resourcing and material support. 23

MCLACHLAN

CONCLUSION

This chapter has argued that there is a range of factors that need to be considered in relation to the early years curriculum and literacy policy for the countries in Africa, and that avoiding the ideological traps that have constrained literacy policy in other countries is important (Olsen, 2009). Any literacy policy needs to be based on two things: on relevant language and literacy practices in African countries and on the evidence generated by research regarding the knowledge and skills that children need for literacy acquisition. One fundamental policy issue to be considered is that children are highly likely to enter school across many of the African countries with few early childhood experiences, or with experiences resulting from a focus on pediatric care, and with limited literate cultural capital (Tunmer and Chapman, 2015). Although children will have had language and literacy experiences in their homes and communities, they may not transition well to a standardized early literacy curriculum. For this reason, literacy policymakers need to consider Olsen’s (2009) policy challenge of ‘Read what’ and ‘How well’ and develop policy that will cater to the needs of diverse multilingual learners. This chapter has argued that a coherent literacy policy should build on the work of Awopegba, Oduolowu, and Nsamenang (2013) and their principles for learning in the early years in the countries of Africa, as well as draw on the research on the benefits of literacy-rich curriculum and evidence from cognitive science and neurolinguistic research that shows that children need to learn the predictors of literacy acquisition, such as alphabet knowledge, phonological awareness, and an extensive vocabulary (NELP, 2009). Teacher knowledge, teacher education, and ongoing professional learning are important factors in the success of an early reading curriculum, because providing differentiated early reading instruction is imperative when dealing with multilingual learners with diverse early literacy experiences (McLachlan and Arrow, 2015). A strong initial teacher education programme is called for, as is increasing the resourcing and infrastructure in schools. A standardized curriculum is unlikely to work, but a literacy policy that advocates for a differentiated curriculum that supports a range of learners and developmental levels is more likely to succeed. Approaches to literacy need to build on community-based approaches to literacy (Awopegba et al., 2013) and to recognize children’s rights to become literate while also capitalizing on research evidence regarding what helps children learn to read. Although some strategies for supporting early literacy and ongoing assessment were proposed, this remains an area of ongoing interest for teachers, researchers, and policymakers. REFERENCES Allington, R. (2010). Recent federal education policy in the United States. In D. Wyse, R. Andrews, & J. V. Hoffman (Eds.), International handbook of educational policy and practice. New York, NY: Routledge.

24

EARLY READING CURRICULUM Al-hajaya, N. (2012). Impact of the systemic approach on literacy achievement of Jordanian 1st graders at Mu’tah University Model School. International Education Studies, 5(1), 100–107. doi:10.5539/ ies.v5n1p100 Arrow, A., McLachlan, C., & Greaney, K. (2015). Teacher knowledge needed for differentiated instruction. In W. E. Tunmer & J. W. Chapman (Eds.), Equity and excellence in literacy education: The case of New Zealand (pp. 171–193). Sydney: Palgrave Macmillan. Ausubel, D., Novak, J. D., & Hanesian, H. (1978). Educational psychology: A cognitive view. New York, NY: Holt, Rinehart & Winston. Awopegba, P. O., Oduolowu, E. A., & Nsamenang, A. B. (2013). Indigenous early childhood care and education (IECCE) curriculum framework for Africa: A focus on context and contents. Addis Ababa: UNESCO International Institute for Capacity Building in Africa (IICBA). Boyer, N., & Ehri, L. C. (2011). Contribution of phonemic segmentation instruction with letters and articulation pictures to word reading and spelling in beginners. Scientific Studies of Reading, 15(5), 440–470. Bruner, J. (1960). The process of education. Cambridge, MA: Harvard University Press. Buckingham, D. (2013). Media education: Literacy, learning and contemporary culture. New York, NY: Wiley. Cleghorn, A., & Prochner, L. (2010). Shades of globalisation: Views from India, South Africa and Canada. Rotterdam: Sense. Connor, C. M., Morrison, F. J., & Katch, L. E. (2004). Beyond the reading wars: Exploring the effect of child-instruction interactions on growth in early reading. Scientific Studies in Reading, 8, 305–336. Cunningham, A. E., Perry, K. E., Stanovich, K. E., & Stanovich, P. J. (2004). Disciplinary knowledge of K-3 teachers and their knowledge calibration in the domain of early literacy. Annals of Dyslexia, 54(1), 139–167. Department of Education (2014). Statutory framework for the early years foundation stage: Setting the standards for learning, development and care for children from birth to five. http://www.foundationyears.org.uk/files/2014/07/EYFS_framework_from_1_September_ 2014__with_clarification_note.pdf Ehri, L. C. (2005). Learning to read words: Theory, findings, and issues. Scientific Studies in Reading, 9, 167–188. Evans, M. A., & Shaw, D. (2008). Home grown for reading: Parental contributions to young children’s emergent literacy and word recognition. Canadian Psychology, 49, 89–95. Friere, P., & Macedo, D. (1987). Literacy: Reading the word and the world. South Hadley, MA: Bergin & Garvey. Garcia, M., Pence, A., & Evans, J. (Eds.) (2008). Africa’s future, Africa’s challenge: Early childhood care and development in sub-Saharan Africa. Washington, DC: World Bank. Gay, G. (2010). Culturally responsive teaching: Theory, research and practice. New York, NY: Teachers’ College Press. Heath, S. B. (1983). Ways with words: Language, life, and work in communities and classrooms. Cambridge, UK: Cambridge University Press. Ifakachukwu, O. (2011). Early childhood education: An overview. International NGO Journal, 6(1), 30–34. Imoh, A. T., & Ansell, N. (2013). Children’s lives in an era of children’s rights: The progress of the Convention on the Rights of the Child in Africa. New York, NY: Routledge. Kagitcibasi, C. (2007). Family, self and human development across cultures: Theory and applications (2nd ed.). Mahwah, NJ: Lawrence Erlbaum. Kalantzis, M., & Cope, B. (2012). Literacies. Port Melbourne, Australia: Cambridge University Press. Knobel, M., & Lankshear, C. (2003). Researching young children’s out-of-school literacy practices. In N. Hall, J. Larson, & J. Marsh (Eds.), Handbook of early childhood literacy (pp. 51–65). London: Sage. Li, H., & Rao, N. (2005). Curricular and instructional influences on early literacy attainment: Evidence from Beijing, Hong Kong and Singapore. International Journal of Early Years Education, 13(3), 235–253.

25

MCLACHLAN Makin, L., Jones Díaz, C., & McLachlan, C. (Eds.) (2007). Literacies in childhood: Changing views, challenging practice. Marrickville, Australia: Elsevier. McLachlan, C., & Arrow, A. (2011). Literacy in the early years in New Zealand: Policies, politics and pressing reasons for change. Literacy, 45(3), 126–133. McLachlan, C., & Arrow, A. (2015). Literacy and the early education curriculum in New Zealand. In W. E. Tunmer & J. W. Chapman (Eds.), Equity and excellence in literacy education: The case of New Zealand. Sydney: Palgrave Macmillan. McLachlan, C., Edwards, S., Margrain, V., & McLean, K. (2013). Children’s learning and development: Contemporary assessment in the early years. Sydney: Palgrave MacMillan. McLachlan, C., Nicholson, T., Feilding-Barnsley, R., Mercer, L., & Ohi, S. (2012). Literacy in early childhood and primary: Issues, challenges, solutions. Melbourne: Cambridge University Press. Ministry of Education (2007). The New Zealand curriculum. Wellington, New Zealand: Learning Media. Ministry of Education (2012). Literacy learning progressions. Wellington, New Zealand: Learning Media. http://www.literacyprogressions.tki.org.nz/ Moats, L. C. (1994). The missing foundation in teacher education: Knowledge of the structure of spoken and written language. Annals of Dyslexia, 44, 81–102. Moats, L. C. (2000). Speech to print: Language essentials for teachers. Baltimore, MD: Paul H. Brookes. Moats, L. C. (2009). Knowledge foundations for teaching reading and spelling. Reading and Writing: An Interdisciplinary Journal, 22, 379–399. Moll, L., Amanti, C., Neff, D., & Gonzalez, N. (1992). Funds of knowledge for teaching: Using a qualitative approach to connect homes and classrooms. Theory into Practice, 31(2), 132–141. Morrow, L. M. (2009). Literacy development in the early years: Helping children to read and write (6th ed.). Boston, MA: Pearson. Muter, V. (1994). Influence of phonological awareness and letter knowledge on beginning reading and spelling development. In C. Hulme & M. Snowling (Eds.), Reading development and dyslexia (pp. 45–62). London: Whurr. NELP [National Early Literacy Panel] (2009). Developing early literacy: Report of the National Early Literacy Panel. Washington, DC: National Institute for Literacy. New London Group (1996). A pedagogy of multiliteracies: Designing for social futures. Harvard Educational Review, 66(1), 60–92. Nicholson, T. (2005). At the cutting edge: The importance of phonemic awareness in learning to read and spell. Wellington, New Zealand: NZCER Press. Olson, D. R. (2009). Literacy, literacy policy, and the school. In D. Olson & N. Torrance (Eds.), The Cambridge handbook of literacy (pp. 566–576). Cambridge, UK: Cambridge University Press. Pence, A., & Ashton, E. (2016). Early childhood research in Africa: The need for a chorus of voices. In A. Farrell, S. Kagan, & E. Tisdall (Eds.), The Sage handbook of early childhood research (pp. 380–397). London: Sage. Phillips, B. M., Clancy-Menchetti, J., & Lonigan, C. (2008). Successful phonological awareness instruction with preschool children. Topics in Early Childhood Special Education, 28(1), 3–17. Phillips, B. M., & Lonigan, C. J. (2009). Variations in the home literacy environment of preschool children: A cluster analytic approach. Scientific Studies in Reading, 13, 146–174. Piasta, S. B., Connor, C. M., Fishman, B. J., & Morrison, F. J. (2009). Teachers’ knowledge of literacy concepts, classroom practices and student reading growth. Scientific Studies in Reading, 13, 224–248. Sénéchal, M. (2011). Relations between home literacy and child outcomes. In S. B. Neuman & D. K. Dickinson (Eds.), Handbook of early literacy research: Volume 3. (pp. 175–188). New York, NY: Guilford Press. Spear-Swerling, L., & Zibulsky, J. (2014). Making time for literacy: Teacher knowledge and time allocation for instructional planning. Reading and Writing, 27, 1353–1378. Spencer, T. G., Falchi, L., & Ghiso, M. P. (2011). Linguistically diverse children and educators (re)forming early literacy policy. Early Childhood Education Journal, 39, 115–123. Stahl, S. A., & Yaden, D. B. (2004). The development of literacy in preschool and primary grades: Work by the Centre for the Improvement of Early Reading Achievement. Elementary School Journal, 105(2), 141–165.

26

EARLY READING CURRICULUM Stanovich, K. (2000). Progress in understanding reading: Scientific foundations and new frontiers. New York, NY: Guilford Press. Tabors, O., & Snow, C. (2001). Young bilingual children and early literacy development. In S. B. Neuman & D. K. Dickison (Eds.), Handbook of early literacy research. New York, NY: Guilford Press. Tagoilelagi-Leota, F., McNaughton, S., MacDonald, S., & Ferry, S. (2005). Bilingual and biliteracy development over the transition to school. International Journal of Bilingual Education and Bilingualism, 8(5), 455–479. Teale, W., Hoffman, J., Paciga, K., Lisy, J. G., Richardson, S., & Berkel, C. (2009). Early literacy: Then and now. In J. V. Hoffman & Y. M. Goodman (Eds.), Changing literacies for changing times: An historical perspective on the future of reading research, public policy and classroom practices (pp. 76–97). London: Routledge Thomson, P. (2002). Schooling the rustbelt kids: Making the difference in changing times. Crows Nest, NSW: Allen & Unwin. Treibel, A. (2005). Literacy in developed and developing countries. In N. Bascia, A. Cumming, A., Datnow, K., Leithwood, & D. Livingstone (Eds.), International handbook of educational policy (pp. 793–812). London: Springer. Tunmer, W. E., & Chapman, J. W. (Eds.) (2015). Excellence and equity in literacy education: The case of New Zealand. Basingstoke: Palgrave Macmillan. Tunmer, W. E., Chapman, J. W., & Prochnow, J. E. (2006). Literate cultural capital at school entry predicts later reading achievement: A seven year longitudinal study. New Zealand Journal of Educational Studies, 41, 183–204. UNESCO BREDA [Bureau régional de l’éducation en Afrique] (2012). Early childhood care and education regional report: Africa. Dakar: UNESCO BREDA. Vygotsky, L. S., Rieber, R. W., & Carton, A. S. (1987). The collected works of L. S. Vygotsky. New York, NY: Plenum Press. Walpole, S., Justice, L. M., & Invernizzi, M. A. (2004). Closing the gap between research and practice: Case study of school-wide literacy reform. Reading and Writing Quarterly, 20, 261–283. Wasik, B. A., Bond, M. A., & Hindman, A. (2006). The effects of a language and literacy intervention on Head Start children and teachers. Journal of Educational Psychology, 98(1), 63–74. Wasik, B. A., & Iannone-Campbell, C. (2012). Developing vocabulary through purposeful, strategic conversations. The Reading Teacher, 66(4), 321–332. Whitehurst, G. J., & Lonigan, C. J. (1998). Child development and emergent literacy. Child Development, 69, 848–872. Whitehurst, G. J., & Lonigan, C. J. (2001). Emergent literacy: Development from prereaders to readers. In S. B. Neuman & D. K. Dickinson (Eds.), Handbook of early literacy research (pp. 11–29). New York, NY: Guilford Press. Wong Fillmore, L., & Snow, C. E. (2000). What teachers need to know about language. Washington, DC: US Department of Education.

27

PETER AFFLERBACH

2. ALIGNING CURRICULUM AND ASSESSMENT IN EARLY READING EDUCATION

INTRODUCTION

Achieving the alignment of curriculum and assessment is a perennial challenge in the field of children’s early reading development. Historically, educational stakeholders and policymakers have sought to achieve this alignment by designing curricula to develop young students’ cognitive strategies and skills, and then designing assessments that measure their growth. In this chapter, I argue that, as critical as these strategies and skills are in early reading development, the narrow focus on these dimensions in alignment efforts is insufficient to create successful, lifelong readers. To be effective, early reading programmes must take into account the findings of recent research that point to the need to work on a wide range of factors involved in children’s reading growth, such as motivation and engagement, self-efficacy, and metacognition. Similarly, the methods used to assess reading should reflect a broader conception of how reading develops. Any discussion of the alignment between curriculum and assessment must, therefore, give due consideration to the complexity of the construct of early reading development. I begin by describing how curriculum and assessment align with constructs and standards. Next, I argue that the construct of early reading development is continually evolving, and that attempts to create optimal curriculum and assessment alignments must take this dynamic nature of the construct into account. The following section then looks at three factors that strongly influence early reading development: motivation and engagement; self-efficacy; and metacognition. These factors, in addition to the traditional focus on cognitive strategies and reading skills, must be considered from both curricular and assessment perspectives. The chapter goes on to discuss types of assessment and curriculum design that reflect this broader view of how students learn. I then focus on the differences and similarities between the intended curricula and the curricula as actually enacted. The chapter concludes with a description of the fundamental principles involved in aligning curriculum and assessment in the field of early reading. THE IMPORTANCE OF ALIGNMENT: CONSTRUCTS, STANDARDS, CURRICULA AND ASSESSMENT

The alignment between curriculum and assessment is merely one of several alignments that contribute to the success of educational efforts. In the field of early reading

© JOINTLY BY IBE-UNESCO AND KONINKLIJKE BRILL NV, LEIDEN, 2019 DOI: 10.1163/9789004402379_003

AFFLERBACH

education, it is just as necessary for curriculum and assessment to align with constructs and standards. These various components to be aligned are represented in Figure 1. Level 1: The construct of ‘early reading’ Level 2: Early reading standards and benchmarks Level 3: Early reading curriculum and instruction Level 4: Assessment of early reading Figure 1. Necessary alignments for curriculum and assessment of early reading

As indicated in Figure 1, several levels are involved in aligning curriculum and assessment. First, the construct of early reading development emerges from educational research and experience (Level 1). Second, based on this construct, standards are developed (Level 2) to define anticipated reading development for most children. Standards equate to realistic goals, and most students are expected to reach these goals given appropriate instruction and opportunities for learning, in keeping with a normal development. Third, curriculum and instruction (Level 3) are designed to reflect the standards, and to help students meet them. Curriculum and instruction are fine-tuned to the developmental trajectory of students’ reading growth so as to be pedagogically effective. Fourth, students undergo assessment (Level 4), which is aligned with the curriculum, standards, and constructs. Summative assessment – assessment that occurs at the end of a defined instructional period to evaluate learning outcomes – is aligned with the early reading curriculum and tells us if students are attaining the expected standards. Formative assessment – assessment that occurs during the learning process with the purpose of providing feedback on student progress – is aligned with daily curriculum goals. The latter type of assessment describes student development ‘in the moment’, and helps teachers provide students with effective instruction on the path towards meeting standards. When this dual approach to assessment is well mapped onto early reading, we have achieved construct validity. It has been argued that construct validity is the most important form of validity, because it focuses on the degree to which an assessment accurately derives from, and portrays, the phenomenon being measured (Messick, 1984, 1989). In the present case, the construct in question is that of early literacy. Alignment is an important factor in achieving construct validity, as well as being essential to all four levels described above. When we connect any of the four components presented in Figure 1, we can analyse their alignment. For example, if we are interested in how assessment aligns with the construct of early reading, we can evaluate our assessments and see if they effectively map onto a robust and up-todate conceptualization of early reading. If we are interested in how well our standards for early reading are represented by the construct, we can map the standards onto the construct to determine their points of convergence. Again, if we are concerned about how assessment relates to curriculum, we can map reading assessment targets 30

ALIGNING CURRICULUM AND ASSESSMENT

to the intended outcomes of early reading instruction. This process has an important practical application, as it helps to optimize the efforts and resources allocated to early reading programmes by identifying programmes’ shortcomings or strengths. Achieving alignment is challenging work, but with the challenge comes the promise of early reading education that effectively addresses students’ reading development needs in all their diversity, as well as the promise of assessment that enables us to make detailed and accurate inferences about student growth in reading. Carefully aligned assessment reflects the breadth and depth of students’ achievements, teachers’ accomplishments, and the efficiency of the instructional strategies (Pellegrino, Chudowsky, and Glaser, 2001; Porter, 2006). CONCEPTUALIZING SUCCESSFUL EARLY READERS

Success in early reading programmes is commonly measured in terms of children’s cognitive growth. This approach to conceptualizing success in early readers is demonstrated in the use of reading assessments that focus on cognitive strategies and skills, and on gains in information in specific areas of content. For example, the PIRLS Assessment Framework draws on the 1991 Reading Literacy Study by the International Association for the Evaluation of Educational Achievement (IEA), which defines reading literacy as ‘the ability to understand and use those written language forms required by society and/or valued by the individual’ (Mullis and Martin, 2015, p. 11). PIRLS and pre-PIRLS, the reading comprehension assessments designed by the IEA, emphasize the importance of using strategies and skills to gain information from texts. They define reading development as the ability to read and comprehend increasingly difficult texts and correctly answer comprehension questions. A sample of reading assessment items from pre-PIRLS typifies this emphasis. For example, after reading a passage about soccer, students answer the following questions (Mullis and Martin, 2015): ‡ ‡ ‡ ‡ ‡

Where was Charlie at the beginning of the story? Who was practicing for a soccer tournament? Write two things that Charlie did during his soccer game. What were Jan’s plans for the summer? What did Charlie keep looking for during the summer?

The above exercise draws principally on cognitive skills and undeniably constitutes a test of reading ability. Nonetheless, reading ability cannot be attributed to cognitive skills alone at the expense of a wide range of other factors that influence how early readers develop. Below I describe three of these factors, namely motivation, selfefficacy, and metacognition, and discuss how they contribute to the development of early reading ability. This is not to ignore that additional factors also contribute to early reading development, such as children’s attributions concerning their own reading performance, that is, children’s degrees of self-esteem or their self-concept. 31

AFFLERBACH

Successful early reading programmes influence not only students’ cognitive development, but also their motivation and engagement (DeNaeghel, Van Keer, Vansteenkiste, and Rosseel, 2012; Schiefele, Schaffner, Moller, and Wigfield, 2012; Zhou, Ma, and Deci, 2009), self-efficacy (Bursuck and Damer, 2014; Ryan and Deci, 2009), and metacognitive development (Baker, 2005; Whitebread et al., 2009). In fact, one may say that the reason these programmes are successful is because, in addition to teaching students reading strategies and skills, they motivate students to read, they foster self-efficacy in how students approach, and work through, reading challenges and they train students to monitor their progress, from the beginning to the completion of a reading task. When the reading curriculum focuses exclusively on reading strategies and skills, it may result in young readers’ excelling in assessments that focus on letter recognition and word pronunciation. Such a curriculum does not, however, directly address students’ needs to develop motivation, self-efficacy, and metacognition. Any assessment aligned on such a curriculum will provide no insight into growth (or lack of growth) in these areas. Consequently, if we value broad outcomes in early reading programmes, we must use assessments that help us understand students’ development in each of these important areas. Since curricula incorporate instructional materials and procedures that help teachers address diverse reading outcomes, assessment must follow suit – tell us about how children become motivated readers, how they grow to believe in themselves as capable readers and how they take responsibility for monitoring their reading, as well as how their reading strategies and skills grow. The aim is to design a reading curriculum that is broadly based, informed by recent research in the field, and aligned with assessment that proficiently measures students’ growth and achievement. VALIDITY OF EARLY READING ASSESSMENT

In the previous section, I argued that there is a need to rethink the construct of early reading development. The nature of this construct as it currently exists can be inferred from those students who successfully develop as early readers. They are efficient at using cognitive strategies and skills to decode and recognize words, to read fluently, to determine the meanings of new words and develop vocabulary, and to construct meaning from text (NICHD, 2000). They use reading to learn about the world around them, to demonstrate learning, and accumulate knowledge in the content areas. Yet, this construct provides an incomplete picture of early reading development. Successful students are also motivated and engaged (Guthrie and Klauda, 2016). They look forward to reading in school. They identify themselves as readers, and they view reading as a way of learning, of being entertained, and of working socially with others. They are confident, and possess the tools to read and succeed in most reading situations (Schunk and Bursuck, 2016). This self-efficacy relates 32

ALIGNING CURRICULUM AND ASSESSMENT

to motivation (readers who believe in their ability to meet reading challenges are motivated), and perseverance when reading presents a challenge. Finally, these readers are metacognitive – they are able to coordinate increasingly complex and demanding reading tasks as they set goals, monitor their progress, and ascertain that they have reached their goals (Veenman, 2016). Metacognition also enables these developing readers to understand the connections between their motivations to read, their self-efficacy, their effort, and their reading outcomes. It is essential that the curricula and assessment used in the field of early reading development reflect this broader conception of the factors that influence students’ successful growth. Pellegrino et al. (2001) describe a general model for effective assessment in which assessment correlates closely with the initial construct the assessor wishes to measure. In this model, the construct in question is described in detail and then assessment materials and procedures are designed to map clearly onto this construct. As Pellegrino and colleagues state, when we have confidence in our assessmentto-construct mapping, it increases confidence in the interpretation of assessment results, as well as in the inferences we make about students’ learning. In the field of early reading development, assessment that focuses exclusively on the growth of young children’s cognitive strategies and skills in early reading – a model that continues to dominate many reading curricula and assessments – will have, at best, only partial construct validity. Yet, our construct of early reading is what guides our efforts to teach and assess. Therefore, if we want to achieve a comprehensive understanding of children’s development towards independent, successful and lifelong reading, there must be a rethinking of this construct to include other factors (Afflerbach, Pearson, and Paris, 2012). If we agree that there is more to early reading development than acquiring cognitive strategies and skills, we are, in effect, expanding the construct of early reading. This has major implications for curriculum and assessment. We do not expect young children to be expert readers, but reading curriculum and assessment for young children should focus on how well (and whether) the broad foundation for successful, lifelong reading is being set. WHAT OTHER FACTORS MATTER IN EARLY READING DEVELOPMENT?

Stanovich (1986) investigated the phenomenon whereby some early readers develop at exponential rates, and he reviewed studies of elementary students’ cognitive development. He proposed the ‘Matthew Effect hypothesis’, which posits that early success with reading begets subsequent success, as both reading comprehension and vocabulary grow. Based on this hypothesis, I would like to propose two related ideas. First, the Matthew Effect is not restricted to cognitive strategies and skills. Research shows that students’ positive or negative learning outcomes in classrooms influence other aspects of development (Guthrie, Wigfield, and You, 2012). A series of successes as young children begin their reading careers results in greater motivation for reading, increased engagement with reading, enhanced self-efficacy as a reader, 33

AFFLERBACH

and habits of metacognition. Second, we must also be aware of negative Matthew Effects – the accumulation of early negative experiences with reading leads students to avoid reading, which entails diminished self-efficacy and a lack of motivation and engagement with reading. Stanovich (1986) attributes initial and ongoing lack of success at reading to individual differences in cognitive skills. This lack of success can lead students to a critical juncture, where they struggle to construct meaning and have little or no inclination to try to read: Readers of differing skill soon diverge in the amount of practice they receive at reading and writing activities. They also have different histories of success, failure, and reward in the context of academic tasks. The long-term effects of such differing histories could act to create other cognitive and behavioural differences between readers of varying skill …. There is already some evidence suggesting that differences in self-esteem, rather than being the cause of achievement variability, are actually consequences of ability and achievement. (p. 373) The excerpt suggests that there may be complex interactions between early readers’ developmental processes that exert further influence on reading progress. Advocating for change in how we characterize early reading development requires evidence that supports our advocacy. Fortunately, ongoing research provides this evidence, demonstrating what accomplished reading teachers already know: readers’ motivation, self-efficacy, and metacognition play central roles in reading development (Afflerbach, 2016). In the following sections, I provide a brief overview of findings in these three areas that support the claim that early reading development depends on more than strategy and skill, and that curriculum and assessment must evolve to espouse a broader view of educational outcomes. MOTIVATION AND ENGAGEMENT IN EARLY READING

Motivation influences students’ daily and long-term approaches to reading (McRae and Guthrie, 2009). Students with histories of success of reading in school tend to be motivated to read more, while students who experience early and regular failure in reading detach from reading. Accordingly, motivation acts to engage or disengage students. The engaged student has more opportunities to experience and practice reading, more opportunities to learn new vocabulary, and more opportunities to discuss what is read with classmates and family. The disengaged student, in contrast, avoids reading and thus loses opportunities to practice and to continue to learn to be a better reader. Studies have shown that reading engagement is fuelled both by motivational processes and cognitive strategies during reading (Guthrie, Coddington, and Wigfield, 2009), and the effect of reading instruction on reading achievement is mediated by students’ motivation (Guthrie et al., 2012). In short, the results 34

ALIGNING CURRICULUM AND ASSESSMENT

of teaching are enhanced when students are engaged. Students’ early reading competence is acquired by articulating knowledge of reading and text, text processing skills and strategies, with personal commitment and motivation (Alexander, 2010). Motivational processes influence how children use their existing strategies, skills, and knowledge, how well they acquire new strategies, skills, and knowledge, and how well they transfer these new strategies, skills, and knowledge to situations they have not encountered previously. Thus, students’ learning and use of early reading strategies and skills are appreciably influenced by motivational factors. Engagement exerts a powerful influence on early readers. As noted by the OECD (2014), ‘Engagement in reading can be a consequence, as well as a cause, of higher reading skill, but the evidence suggests that these two factors are mutually reinforcing’ (p. 8). Put another way, readers who are more engaged read more text, and readers who read more text are more engaged. The influence of engagement has been found to help readers from homes and communities marked by low socio-economic status. Cummins (2016) states that ‘the negative effects of socioeconomic disadvantage can be “pushed back” in schools and classrooms where students have access to a rich print environment and become actively engaged with literacy’ (p. 231). SELF-EFFICACY IN EARLY READING

As early readers learn and use strategies and skills, they must also develop the belief that they can succeed at reading. Young readers’ self-efficacy is central to this belief, and to related academic performance (Lee and Jonson-Reid, 2016). Bandura (2006) powerfully expresses this idea: Among the mechanisms of human agency, none is more central or pervasive than belief of personal efficacy. Unless people believe they can produce desired effects by their actions, they have little incentive to act, or to persevere in the face of difficulties [emphasis added]. Whatever other factors serve as guides and motivators, they are rooted in the core belief that one has the power to effect changes by one’s actions. (p. 308) Just as instruction needs to help students to develop reading strategies and skills, it must also ensure that students develop ‘belief in self’ as successful readers. Research has demonstrated that higher self-efficacy is a factor in enhanced classroom achievement (Solheim, 2011). Students with high self-efficacy make fewer attributions of performance to external causes, such as luck, task difficulty, and teacher help (Shell, Colvin, and Bruning, 1995). Furthermore, highly efficient students are more strategic (Li and Wang, 2010). Students with high self-efficacy share the following characteristics: they have high aspirations for personal success, they sustain effort when facing learning challenges, they view problems as challenges to be overcome, and they equate poor performance with insufficient effort. In contrast, students with low self-efficacy 35

AFFLERBACH

share low aspirations for personal success, they have a limited commitment to goals, they equate poor performance with poor capability, and they are less likely to take on, and persevere with, challenging tasks. These students attribute school success to external factors, including luck and ease of task. Students with low self-efficacy are more vulnerable to stress and depression. The implications of the research are clear: students whose initial and ongoing early reading experiences do not support the development of self-efficacy are at risk of failure (Dweck, 1986; Schunk and Bursuck, 2016). METACOGNITION IN EARLY READING

Hennessey (1999) characterizes metacognition as follows: Awareness of one’s own thinking, awareness of the content of one’s conceptions, an active monitoring of one’s cognitive processes, an attempt to regulate one’s cognitive processes in relation to further learning, and an application of a set of heuristics as an effective device for helping people organize their methods of attack on problems in general. (p. 3) Where reading is concerned, as students grow metacognitively, they can increasingly initiate, work through, and successfully complete reading tasks on their own. However, early readers are not equipped with these skills; so it is imperative that curricula model, and teach, the critical mindsets and strategies required to develop them (Bryce and Whitebread, 2012; Borkowski and Turner, 1990; Veenman, van Hout-Wolters, and Afflerbach, 2006). As students’ metacognitive skills develop, they set goals, monitor their understanding, are sensitive to disruptions in meaning, determine when and where there are problems, solve the problems, and get back on track. During and after reading, metacognitive readers regularly judge their progress towards their reading goals. Research shows that metacognition promotes academic learning (Paris and Winograd, 1990), that struggling readers can learn comprehension-monitoring strategies (Palincsar and Brown, 1984), and that metacognitive strategies improve reading comprehension (Silven, 1992). Self-regulated learning, a form of metacognition, involves not only cognitive processes as learning takes place, but also affect, insofar as it sustains learning processes (Greene and Azevedo, 2007). This finding is supported by research that shows that metacognition involves affective, motivational, and behavioural components that provide learners with the capacity to adjust their actions and goals to achieve desired results (Boekaerts, Pintrich, and Zeidner, 2000). As with motivation, engagement, and self-efficacy, it is important that early reading curricula provide a strong foundation in metacognition for the growth of early readers. We do not expect our early readers to fluently read complex texts and then complete complex metacognitive tasks, but we provide them with initial skills and strategies that put them squarely on the path to doing so. 36

ALIGNING CURRICULUM AND ASSESSMENT

In short, student development is influenced by factors that act both individually and in relation to one another to impact early reading. Given the effect that motivation, self-efficacy, and metacognition have on early reading development, it is important to include them in the conceptualization of students’ reading development. In turn, this conceptualization must be reflected in the design of curriculum and of assessment. Any reading education programme that fails to consider these powerful factors is simply not up to date with current research, and its associated assessments will be only partially aligned with the construct of reading development. The result will be classrooms in which students do not receive attention to all their reading development needs. ALIGNMENT, AUDIENCE AND PURPOSE: THE USEFULNESS OF EARLY READING ASSESSMENT INFORMATION

Having considered the alignment of curriculum and assessment from the perspective of a broad conceptualization of early reading development, I now turn my attention to other types of alignment that influence early reading development. These include the alignment of the audiences and purposes of assessment, the specific types of early reading assessment, and the resources available for assessment. Typically, the audiences of reading assessment are teachers, students, administrators, parents, politicians, and policymakers, whereas the purposes of assessment are variously to measure students’ progress and achievements, to gauge the effectiveness of early reading programmes, and to evaluate school and teacher accountability. Audiences and purposes often find themselves in competition for the limited resources allocated to assessment. For example, when a national education system, regional education authority or school district focuses on summative, high-stakes tests, this approach caters to the needs of those who use the test results (policymakers, politicians, and administrators). These test scores enable them to compare schools and compare students. The diagnostic value of high-stake tests is, however, quite limited: summative test scores provide teachers and students little. They do not get the assessment information best suited to improving daily instruction and student performance. If assessment information does not adequately reflect students’ immediate instructional needs (and thereby inform teaching and learning), test scores are less likely to improve. It is worth noting, further, that in the above scenario, the information provided by tests is restricted to reading strategies and skills. Thus, this scenario also serves as a reminder that achieving alignment between assessment and all legitimate audiences and purposes is a crucial step in enhancing students’ reading ability. In the above example, assessment is aligned with the needs of specific audiences (policymakers and administrators) while neglecting others’ (teachers and students). In short, the example illustrates how political factors may weigh on alignment efforts, because policymakers and legislators may have different views from teachers as to what constitutes effective early literacy instruction. These views may (or may 37

AFFLERBACH

not) be informed by research (Russell, 2011; Spillane, Reiser, and Reimer, 2002). Similarly, the alignment of curriculum and assessment is influenced by economic factors, as few nations or districts possess unlimited assessment budgets. If resources are largely directed towards summative, high-stakes tests, fewer resources will be available for the formative assessments vital for teachers to accurately understand young readers’ development and to plan lessons. An early reading assessment budget used up by costs of purchasing, practicing, administering, scoring, and reporting tests will also leave fewer resources for the professional development of teachers’ formative classroom-based assessments. Another necessary consideration, in addition to opportunity cost, is how each part of the assessment programme provides useful information to a specific audience. Taking this into account is an important step in the development of early reading programmes. If the review of existing and planned assessments reveals misalignment between assessments, audiences, and purposes, then this misalignment must be addressed. Audiences, like purposes, can be aligned with either formative or summative assessment. Mansell et al. (2009) remind us of the important differences between the two. Summative assessment occurs at the end of learning periods, while formative assessment occurs while teaching and learning is ongoing. The former is intended to assess students’ knowledge and understanding at a fixed point in time; the latter, to help teachers best develop students’ knowledge. Effective early reading assessment programmes are aligned with the curriculum. Effective assessment programmes also reflect an internal logic – one that represents a strategic combination of formative and summative assessment. It is imperative to align reading assessments not only with curricula, but also with the needs of those who use the assessment: teachers and students, as well as administrators and policymakers. REPRESENTATIVE ASSESSMENTS AND CURRICULUM

Accurately assessing early reading development is key to effective teaching and student success. Early reading instruction abounds with tests, quizzes, and skill sheets that provide feedback on students’ strategy and skill development. There is a lack, however, of comprehensive testing that reflects the place of motivation, self-efficacy, and metacognition in reading development. Specific assessments have been developed to focus on each of these components, such as on metacognition (Metacognitive Awareness of Reading Strategies Inventory/MARSI) (Mokhtari and Reichard, 2002), and attitudes towards reading (McKenna and Kear, 1990). Similarly, students’ motivation to read can be assessed using the Motivation to Read Profile (MRP) (Gambrell, Palmer, Codling, and Mazzoni, 1996), while students’ self-concepts and self-efficacy can be evaluated with the Reading Self-Concept Scale (RSCS) (Chapman and Tunmer, 1995). It should be noted that the above assessments, while designed for elementary school levels, can prove difficult to administer with some young students, as they require a certain degree of linguistic comprehension. 38

ALIGNING CURRICULUM AND ASSESSMENT

This is a challenge going forward. Let us also note that these assessments vary as to the strength of their psychometric underpinnings. While it is important to be aware of these shortcomings, the very fact that these assessments exist is testament to the legitimacy and importance of the constructs they represent. In order to harness the potential they offer for refining our understanding of early reading development, the next steps are to prove their validity and reliability, and then to make the content of each assessment widely accessible to students – in all countries and classrooms. As noted earlier, a considerable number of early reading programmes have a strong (if not an exclusive) focus on the teaching and learning of the mechanics of reading. These programmes are comprised of curricula that incorporate strategies and skills based on phonemic awareness, phonics, and fluency. They may also focus on young children’s concepts of print. The focus of these mechanics-oriented programmes is reflected in the assessments that accompany them, including Dynamic Indicators of Basic Early Literacy Skills (DIBELS) (Center on Teaching and Learning, 2016) and Early Grade Reading Assessment (EGRA) (Gove and Wetterberg, 2011). For an early reading curriculum to be close aligned with either EGRA or DIBELS may indicate that its students are learning essential strategies and skills related to phonemic awareness, phonics, and fluency. It would be a mistake, however, to interpret a high degree of alignment between an early reading curriculum and EGRA as proof that the needs of young developing readers are being met. These assessments are suitable for measuring the mechanical aspects of early reading development (recognizing letters of the alphabet, reading simple words, understanding sentences and paragraphs, and listening with comprehension), but they ignore the other important aspects of early reading development highlighted in this chapter, namely motivation, engagement, self-efficacy, and metacognition. Few examples exist in which reading curriculum and assessment are developed concurrently, or in which the designs of reading curriculum and assessment encompass more than just reading strategy and skill development. One such example, however, is Concept-Oriented Reading Instruction (CORI) (Guthrie, 2016), which was designed to also take into account student motivation and engagement. CORI seeks to combine student motivation, conceptual knowledge in science, and reading strategy instruction. A core tenet of CORI is that engagement and motivation are essential to the development of literacy learners. As part of the curricular goals set by CORI, assessment is designed to reflect the broad focus on cognitive, affective, and conative outcomes. This assessment is based on a series of measures created by CORI: the Reading Activity Inventory; the Reading Engagement Index; the Motivations for Reading Questionnaire; and Young Reader Motivation Questionnaires. The CORI approach, including curriculum and assessment materials, serves as a model for conceptualizing student development more broadly. Furthermore, the process by which CORI was developed, interweaving curriculum and assessment development so that they become intertwined, provides an interesting model for how to achieve curricular alignment. 39

AFFLERBACH

Finally, where assessment is concerned, let us not forget the essential role played by the teacher in the classroom. Teachers are uniquely situated to enact curriculum and conduct assessment of the broad range of developing-reader characteristics. Using standardized observation schedules, teachers can identify students who are engaged and motivated or disengaged and unmotivated, believing in themselves and applying effort or doubtful of their efficacy and avoiding work. Teachers can also use questioning routines to determine what students find motivating and engaging, students’ self-concepts as learners, and students’ beliefs about their abilities (Afflerbach, 2012). The challenge here is to provide professional development for teachers so that they can become proficient in such classroom assessment routines. If we neglect to help teachers develop assessment competence and expertise, we will continue to be dependent exclusively on tests, quizzes and worksheets. Continuing professional training in the field of early reading programmes will help teachers become reliable assessors of student development – from strategy and skill to motivation and self-efficacy. THE PROCESS OF ALIGNING EARLY READING PROGRAMME CURRICULUM AND ASSESSMENT

When discussing the importance of alignment above, I argued that the construct of early reading development needs to be redefined to take into account the findings of recent research. I noted that the broad view of early reading development that emerges from this research stands in contrast to the consistently narrow representation of that construct in many early reading curricula; and that narrow representation in turn induces a narrow approach to early reading assessment. The alignment is poor between the construct of early reading, how we teach it and how we assess it. Curricula and assessment must evolve to better reflect the complexity of learning to read and the outcomes of reading, so that they can better contribute to the goal of young children’s becoming successful, enthusiastic readers. Wherever we are in this evolution, determining the alignment of what is being taught with what is being assessed remains paramount. Many efforts to achieve educational alignment have focused on the relationship between standards and assessment (Bhola, Impara, and Buckendahl, 2003; Hermann, Webb, and Zuniga, 2007; Wixson, Fisk, Dutro, and McDaniel, 2002). Aligning curricula and assessment is arguably just as vital (Porter, 2002), although this approach is less frequently adopted in practice. One way of achieving this alignment is by systematically mapping assessment onto curriculum (UNESCO IBE, 2016a). Typically, curriculum and assessment share cognitive targets, or a specific learning outcome expected of students. For example, we can specify early reading curriculum goals, such as understanding spoken stories, building fluency, or identifying and using consonant blends. With these goals in mind, the curriculum that is developed includes syntactically and semantically simple stories for the teacher to read, and consonant blend exercises. These are embedded in teaching sequences that help 40

ALIGNING CURRICULUM AND ASSESSMENT

students learn both of these essential early reading skills. This content is distributed, in lesson format, over a specific time period, during which new material is introduced, taught, practiced and learned. Concurrently, assessment is designed to provide both formative and summative information. Summative assessment describes students’ learning: the degree to which early readers comprehend stories that are spoken (or read aloud) by the teacher, and their ability to identify and use consonant blends (for example, /bl-/). Related formative assessment describes student progress in developing these abilities in both areas. Early reading curricula must incorporate goals that are related to, but go beyond, cognitive strategies and skills. Early readers must have positive motivations, high self-efficacy, and metacognitive mindsets to develop into lifelong, successful readers. We can reasonably assume that the developers of early reading curricula have this hope for students. Most early reading curricula, however, fail to directly address these vital factors. In light of this shortcoming, the cognitive goals that underlie early reading curricula ought to be augmented by curricula that also set motivation as a goal, by framing early readers’ motivation as an instructional outcome and incorporating materials that are motivating to read. Teacher support during reading lessons can provide students with additional motivation, as can routines that help students focus on how their reading is improving across the school year. As motivation becomes a featured and expected outcome of early reading curriculum, assessments should follow suit and monitor its development. Motivation profiles and surveys, including the Elementary Reading Attitude Survey (ERAS) (McKenna and Kear, 1990), and the Motivation to Read Profile (MRP) (Gambrell et al., 1996) demonstrate the necessary and positive shift in alignment. A similar approach should be taken to self-efficacy and metacognition. While it is important to achieve alignment between assessment and curricula, we must be clear about which curriculum we are referring to: intended or enacted curriculum? The intended curriculum is that which is represented in learning materials and lesson plans. This is the curriculum developed by curriculum experts. The intended curriculum is the result of planning and design conducted prior to the classroom work done by teachers and students. It is a blueprint for teaching and learning created with the hope that all will go according to plan (UNESCO IBE, 2016b). In contrast, the enacted curriculum represents what actually happens in classrooms (Porter, 2006). The enacted curriculum is the intended curriculum as it is influenced by the demands of schooling: student-teacher ratios, school resources, student diversity, varying degrees of teacher expertise, scheduling norms and quirks, student behaviour, the pressures of high-stakes testing, and so on. The intended curriculum is aspirational – what could happen when planned curricular materials are used in optimal classroom settings. While achieving alignment between assessment and the intended curriculum is an important goal, the alignment between assessment and the enacted curriculum is more vital. Let us consider, for instance, the case in which schools using a broadly conceptualized early reading curriculum (that is, one that incorporates motivation and self-efficacy 41

AFFLERBACH

along with strategies and skills) are pressured to take a high-stakes test. In this case, the intended curriculum reflects what we know to be important in the development of early readers, while the enacted curriculum reflects the efforts that have gone into preparing to take a specific test. Thus, the alignment of curriculum and assessment in this case may reveal that the use of high-stakes assessments changes curriculum from intended to enacted. When assessment is closely aligned with either the intended or the enacted curriculum, it is poorly aligned with the other; so those who are tasked with aligning curriculum and assessment must be clear about which type of curriculum is the focus. The intended curriculum for a school, district or nation can be ascertained by examining and analysing existing curricular materials in conjunction with documentation on educational goals and learning targets. In contrast, determining what constitutes enacted curriculum requires the collection of data that describes what happens during the school day in classrooms. Formal instruments for conducting this important work include the Surveys of Enacted Curriculum (SEC) (Blank, 2004), which are used with teachers to gather information about instructional practice and the nature of the learning content that students encounter in classrooms. The creators of SEC report that it can be used to analyse alignment between current instruction and learning standards and assessments. However, if they fail to specify which curriculum (intended or enacted) is being investigated for alignment with assessment, the findings of studies on curriculum-assessment alignment are of questionable use. CONCLUSIONS

Aligning curriculum and assessment plays a critical role in optimizing our efforts to foster the early reading development of all students. Achieving alignment is always a complex and demanding process, all the more so as we acknowledge the complexity of early reading development. Successful early readers are competent in using cognitive strategies and skills, and this aspect of reading development is strongly supported by the majority of early reading curricula. However, several other factors influence early readers, in particular motivation, engagement, self-efficacy, and metacognition. To be effective, reading curricula need to also foster these aspects of student development. Once curricula are designed to better represent the complex construct of early reading development, assessment must follow suit: it must also be designed to support teachers’ daily evaluations of student progress, and, conversely, help students achieve long-term goals and standards. A further, necessary alignment is that between assessment audiences and purposes. Policymakers use summative test scores to make policy decisions, and teachers use formative assessment information gained from classroom observation and teacher questions to improve their instruction. It is important to closely monitor this alignment between audience and purpose to optimize learning outcomes. Formative assessment is essential for fostering students’ ongoing development, 42

ALIGNING CURRICULUM AND ASSESSMENT

which summative assessment can then validate. As curriculum and assessment budgets and resources are always limited, careful alignment can ensure that one type of assessment is not used at the expense of the other. In addition, it is necessary to distinguish between an intended and an enacted curriculum. The actual early reading instruction that students experience may well vary from a curriculum published on the subject. Studies must specify which type of curriculum, enacted or intended, assessment is aligned with, and to what degree. Finally, it is imperative to examine all the alignments that contribute to successful early reading programmes. Construct, standards, curriculum, and assessment all require close scrutiny as we determine their individual integrity and their beneficial alignments. REFERENCES Afflerbach, P. (2012). Understanding and using reading assessment, K-12 (2nd ed.). Newark, DE: International Reading Association. Afflerbach, P. (2016). Handbook of individual differences in reading: Reader, text, and context. New York, NY: Routledge. Afflerbach, P., Pearson, D., & Paris, S. (2012). Une clarification théorique: Clarifier les différences entre les compétences en lecture et les stratégies de lecture. Caractères, 44, 17–36. Alexander, P. (2010). Reading into the future: Competence for the 21st century. Educational Psychologist, 47, 259–280. Baker, L. (2005). Developmental differences in metacognition: Implications for metacognitively oriented reading instruction. In S. Israel, C. Block, K. Bauserman, & K. Kinnucan-Welsch (Eds.), Metacognition in literacy learning (pp. 641–79). Mahwah, NJ: Lawrence Erlbaum. Bandura, A. (2006). Guide for constructing self-efficacy scales. In F. Pajares & T. Urdan (Eds.), Selfefficacy beliefs of adolescents (pp. 307–337). Charlotte, NC: IAP. Bhola, D., Impara, J., & Buckendahl, C. (2003). Aligning tests with states’ content standards: Methods and issues. Educational Measurement: Issues and Practices, 22(3), 21–29. Blank, R. (2004). Data on enacted curriculum study: Summary of findings. Washington, DC: Council of Chief State School Officers. Boekaerts, M., Pintrich, P., & Zeidner, M. (Eds.) (2000). Handbook of self-regulation. San Diego, CA: Academic Press. Borkowski, J., & Turner, L. (1990). Transsituational characteristics of metacognition. In W. Schneider (Ed.), Interactions among aptitudes, strategies, and knowledge in cognitive performance (pp. 159– 176). New York, NY: Springer. Bryce, D., & Whitebread, D. (2012). The development of metacognitive skills: Evidence from observational analysis of young children’s behavior during problem-solving. Metacognition and Learning, 7(3), 197–127. Bursuck, W., & Damer, M. (2014). Teaching reading to students who are at-risk or have disabilities (3rd ed.). Boston: Pearson Education. Center on Teaching and Learning (2016). Dynamic indicators of basic early literacy skills. Eugene, OR: University of Oregon, Center on Teaching and Learning. Chapman, J., & Tunmer, W. (1995). Development of young children’s reading self-concepts: An examination of emerging subcomponents and their relationship with reading achievement. Journal of Educational Psychology, 87, 154–167. Cummins, J. (2016). Language differences that influence reading development: Instructional implications of alternative interpretations of the research evidence. In P. Afflerbach (Ed.), Handbook of individual differences in reading: Reader, text, and context (pp. 223–244). New York, NY: Routledge. DeNaeghel, J., Van Keer, H., Vansteenkiste, M., & Rosseel, Y. (2012). The relation between elementary students’ recreational and academic reading motivation, reading frequency, engagement and

43

AFFLERBACH comprehension: A self-determination theory perspective. Journal of Educational Psychology, 104, 1006–1021. Dweck, C. (1986). Motivational processes affecting learning. American Psychologist, 41, 1040–1048. Gambrell, L., Palmer, B., Codling, R., & Mazzoni, S. (1996). Assessing motivation to read. The Reading Teacher, 49, 518–533. Gove, A., & Wetterberg, A. (2011). The early grade reading assessment: Applications and interventions to improve basic literacy. Research Triangle, NC: RTI Press. Green, J., & Azevedo, R. (2007). A theoretical review of Winne and Hadwin’s model of self-regulated learning: New perspectives and directions. Review of Educational Research, 77, 334–372. Guthrie, J. (2016). Concept-oriented reading instruction. College Park, MD: University of Maryland. Guthrie, J., Coddington, C., & Wigfield, A. (2009). Profiles of motivation for reading among African American and Caucasian students. Journal of Literacy Research, 41, 317–353. Guthrie, J., & Klauda, S. (2016). Engagement and motivational processes in reading. In P. Afflerbach (Ed.), Handbook of individual differences in reading: Reader, text, and context. New York, NY: Routledge. Guthrie, J., Wigfield, A., & You, W. (2012). Instructional contexts for engagement and achievement in reading. In S. Christensen, A. Reschly, & C. Wylie (Eds.), Handbook of research on student engagement. New York, NY: Springer. Hennessey, M. (1999). Probing the dimensions of metacognition: Implications for conceptual change in teaching and learning. Paper presented at the annual meeting of the National Association for Research in Science Teaching, Boston, MA, 28–31 March, 1999. Herman, J., Webb, N., & Zuniga, S. (2007). Measurement issues in the alignment of standards and assessments: A case study. Applied Measurement in Education, 20, 101–126. Lee, S., & Jonson-Reid, M. (2016). The role of self-efficacy in reading achievement of young children in urban schools. Child and Adolescent Social Work Journal, 33, 79–89. Li, Y., & Wang, C. (2010). An empirical study of reading self-efficacy and the use of reading strategies in the Chinese EFL context. Asian EFL Journal, 12, 144–162. Mansell, W., James, M., & Assessment Reform Group. (2009). Assessment in schools. Fit for purpose? A commentary by the Teaching and Learning Research Programme. London: Economic and Social Research Council, Teaching and Learning Research Programme. McKenna, M., & Kear, D. (1990). Measuring attitude towards reading: A new tool for teachers. The Reading Teacher, 43, 626–639. McRae, A., & Guthrie, J. (2009). Promoting reasons for reading: Teacher practices that impact motivation. In E. H. Hiebert (Ed.), Reading more, reading better (pp. 55–76). New York, NY: Guilford Press. Messick, S. (1984). The psychology of educational measurement. Journal of Educational Measurement, 21, 215–237. Messick, S. (1989). Validity. In R. L. Linn (Ed.), Educational measurement (3rd ed., pp. 13–103). New York, NY: Macmillan. Mokhtari, K., & Reichard, C. (2002). Assessing students’ metacognitive awareness of reading strategies. Journal of Educational Psychology, 94, 249–259. Mullis, I., & Martin, M. (2015). PIRLS 2016 assessment framework (2nd ed.). Chestnut Hill, MA: Boston College, TIMSS and PIRLS International Study Center. NICHD [National Institute of Child Health and Human Development] (2000). Report of the National Reading Panel. Teaching children to read: An evidence-based assessment of the scientific research literature on reading and its implications for reading instruction. Washington, DC: US Government Printing Office. OECD (2014). PISA 2012 results: What students know and can do: Student performance in mathematics, reading and science. Paris: OECD. http://dx.doi.org/10.1787/9789264208780-en Palincsar, A., & Brown, A. (1984). Reciprocal teaching of comprehension-fostering and comprehension monitoring activities. Cognition and Instruction, 1, 117–175. Paris, S., & Winograd, P. (1990). How metacognition can promote academic learning and instruction. In B. Jones & L. Idol (Eds.), Dimensions of thinking and cognitive instruction (pp. 15–51). Hillsdale, NJ: Lawrence Erlbaum Associates.

44

ALIGNING CURRICULUM AND ASSESSMENT Pellegrino, J., Chudowsky, N., & Glaser, R. (2001). Knowing what students know: The science and design of educational assessment. Washington, DC: National Academies Press. Porter, A. (2002). Measuring the content of instruction: Uses in research and practice. Educational Researcher, 31, 3–14. Porter, A. C. (2006). Curriculum assessment. In J. L. Green, G. Camilli, & P. B. Elmore (Eds.), Complementary methods for research in education (3rd ed., pp. 141–160). Washington, DC: American Educational Research. Russell, J. L. (2011). From child’s garden to academic press: The role of shifting institutional logics in redefining kindergarten education. American Educational Research Journal, 48, 236–267. Ryan, R., & Deci, E. (2009). Promoting self-determined school engagement: Motivation, learning, and well-being. In K. R. Wentzel & A. Wigfield (Eds.), Handbook of motivation at school (pp. 171–195). New York, NY: Routledge. Schiefele, U., Schaffner, E., Moller, J., & Wigfield, A. (2012). Dimensions of reading motivation and their relation to reading behavior and competence. Reading Research Quarterly, 47, 427–463. Schunk, D., & Bursuck, W. D. (2016). Self-efficacy, agency, and volition. In P. Afflerbach (Ed.), Handbook of individual differences in reading: Reader, text, and context (pp. 54–66). New York, NY: Routledge. Shell, D., Colvin, C., & Bruning, R. (1995). Self-efficacy, attribution, and outcome expectancy mechanisms in reading and writing achievement: Grade-level and achievement-level differences. Journal of Educational Psychology, 87, 386–398. Silven, M. (1992). The role of metacognition in reading instruction. Scandinavian Journal of Educational Research, 36, 211–221. Solheim, O. (2011). The impact of reading self-efficacy and task value on reading comprehension scores in different item formats. Reading Psychology, 32, 1–27. Spillane, J., Reiser, B., & Reimer, T. (2002). Policy implementation and cognition: Reframing and refocusing implementation research. Review of Educational Research, 72, 387–431. Stanovich, K. (1986). Matthew effects in reading: Some consequences of individual differences in acquisition of literacy. Reading Research Quarterly, 21, 360–407. UNESCO IBE (2016a). What makes a quality curriculum? Geneva: UNESCO IBE. UNESCO IBE (2016b). Training tools for curriculum development: Reaching out to all learners: A resource pack for supporting inclusive education. Geneva: UNESCO IBE. Veenman, M. (2016). Metacognition. In P. Afflerbach (Ed.), Handbook of individual differences in reading: Reader, text, and context (pp. 26–40). New York, NY: Routledge. Veenman, M., van Hout-Wolters, B., & Afflerbach, P. (2006). Metacognition and learning: Conceptual and methodological issues. Metacognition and Learning, 1, 3–14. Whitebread, D., Coltman, P., Pasternak, D., Sangster, C., Grau, V., & Demetriou, D. (2009). The development of two observational tools for assessing metacognition and self-regulated learning in young children. Metacognition and Learning, 4, 63–85. Wixson, K., Fisk, M., Dutro, E., & McDaniel, J. (2002). The alignment of state standards and assessments in elementary reading. Ann Arbor, MI: Center for the Improvement of Early Reading, University of Michigan. Zhou, M., Ma, W., & Deci, E. (2009). The importance of autonomy for rural Chinese children’s motivation for learning. Learning and Individual Differences, 19, 492–498.

45

NIC SPAULL AND ADAIAH LILENSTEIN

3. ASSESSING EARLY LITERACY OUTCOMES IN BURKINA FASO AND SENEGAL Using DHS and PASEC to Combine Access and Quality

INTRODUCTION

The ability to read for meaning and pleasure is arguably the most important skill that children learn in the early years of primary school. Almost all future learning will depend on this fundamental understanding of the relation between print and spoken language. It is therefore not surprising that literacy, built upon a firm foundation of basic reading, is used as one of the primary measures of school efficacy. In West African countries, as in many other developing countries, many children never learn to read at a basic level in the dominant language, French, which is a foreign language for most students. This is because they do not attend school for long enough, and/or because the quality of their schooling is so low that they never breach this critical threshold. The fact that many children do not acquire basic reading and writing skills in the early years of primary school does not prevent them from proceeding to higher grades. This is largely because many countries adhere to a system of social, rather than competency-based, promotion. Moreover, the formal school curriculum proceeds unchanged, assuming that students have mastered skills in previous grades that only a negligible minority have really attained (Pritchett and Beatty, 2015). Education occupies a pre-eminent role in the social and economic development discourses, both in Africa and internationally. In the academic literature, authors such as Hanushek and Woessman (2008), and Goldin and Katz (2009) have used empirical evidence to show the intimate links between educational expansion, educational quality, and economic growth. Looking more broadly, Easterlin (1981) provided the historical narrative of the 25 largest nations in his presidential address to the Economic History Association. He argued that their economic prosperity depended on diffusing knowledge about new production techniques, which was itself predicated on the establishment and expansion of mass schooling. Yet, as we have mentioned in earlier work (Spaull and Taylor, 2015), education has taken centre stage in the global development discourse not only because of its economic significance. There are also widely acknowledged social benefits, including lower fertility (Basu, 2002), improved child health (Currie, 2009), reduced societal violence and improved human rights (Salmi, 2000), promotion of a national – as opposed to a regional or ethnic – identity (Glewwe, 2002), and,

© JOINTLY BY IBE-UNESCO AND KONINKLIJKE BRILL NV, LEIDEN, 2019 DOI: 10.1163/9789004402379_004

SPAULL & LILENSTEIN

finally, increased social cohesion (Heyneman, 2003). Over and above these singular benefits of education, Sen (1999) and Nussbaum (2006) offer broader theories in which education plays a central role in expanding the individuals’ capabilities and freedoms and in enabling them to pursue the sorts of lives they have reason to value. Burkina Faso and Senegal are both francophone West African countries characterized by low levels of economic development and high rates of poverty. The United Nations Development Programme ranks Burkina Faso as 183rd of the 188 countries included, while Senegal is the 170th (UNDP, 2015). While the academic education literature on these two countries is relatively sparse, two recent reviews of their education systems provide a helpful contextual narrative for schooling in Burkina Faso (Dembélé, Somé, and Ouédraogo, 2015a, 2015b) and Senegal (Manion, 2015; Salmon and Dramani, 2015). One salient feature that we find worth noting is the rapid expansion of access to education in the previous two decades. From 2000/01 to 2007/08, enrolment in Senegalese elementary schools grew from 72% to 90%, while enrolment in middle schools grew from 20% to 39% and enrolment in secondary schools grew from 9% to 16% (Manion, 2015). In Burkina Faso, the expansion has been more dramatic. In 2000/01 there were 5,131 schools in the country (614 of which were private) with 17,456 classrooms. These numbers grew to 11,545 schools (2,279 of which were private) with 43,661 classrooms, representing a two- to threefold expansion in schools/classrooms (Dembélé et al., 2015b). With such a rapid expansion of access to education – in Senegal, but particularly in Burkina Faso – there is inevitably an impact on such learning conditions as class sizes, learning and teaching resources, and the recruitment and training of teachers. This impact is most pronounced when the expansion occurs in a country with little economic activity and few national resources, such as Burkina Faso. In these instances, it is imperative to look at both the proportion of children who are in and out of school (access), and, as well, the learning outcomes of those in school (quality); not simply at one or the other. In this chapter, it will become clear that far too few Burkinabe and Senegalese children are learning to read in French by the end of Grade 5, or acquiring emergent literacy skills by the end of Grade 2. Additionally, we highlight the relationship that inequality has with educational access and literacy. With this challenge in mind, we argue that the need for sustained political and financial support for early reading and writing is a developmental imperative. EDUCATION FOR ALL AND THE DISCONNECT BETWEEN ACCESS AND QUALITY

It is precisely because of the social and economic benefits mentioned earlier that most countries have declared education to be a basic human right, enshrined as it is in both the Universal Declaration of Human Rights (United Nations, 1948) and the African Charter on Human and Peoples’ Rights (African Commission on Human and Peoples’ Rights, 1981). In order to operationalize these aspirational goals, the international community has committed itself to a number of large-scale 48

ASSESSING EARLY LITERACY OUTCOMES

initiatives. These started with the two UNESCO World Conference on Education for All initiatives, which were hosted first in Jomtien (Thailand) in 1990 and then in Dakar (Senegal) in 2000. Here all participating countries expressed a commitment to universal primary education. This goal was also affirmed at the Millennium Summit in New York in 2000, where the Millennium Development Goals (MDGs) were set. The MDGs represented the primary agenda for the global development community from 2000 to 2015. All of these movements have been tremendously successful at expanding access to education. Between 1980 and 2010, the average years of schooling for those aged 15 years and over who were living in developing countries increased from 4.3 years to 7.1 years (Barro and Lee, 2013). However, in the late 2000s, a growing number of scholars highlighted the disconnect between schooling (quantity) and learning (quality) in developing countries, showing that additional years of schooling and new competencies acquired were correlated only very loosely, if at all (for some examples, see Filmer, Hasan, and Pritchett, 2006; Hanushek and Woessman, 2008; Majgaard and Mingat, 2012; Pritchett, 2013). In addition, the academic literature was, and largely remains, almost entirely bifurcated. Articles focus either on access to education, looking at enrolment and grade completion, or on the quality of education, primarily analysing cross-national assessments, such as the Trends in International Mathematics and Science Study (TIMSS), the Progress in Reading Literacy Study (PIRLS), and the Programme for International Student Assessment (PISA). Articles do not, however, address access and quality simultaneously. A cursory overview of the literature also reveals that the vast majority of studies looking at education in Africa focus on issues of access to education rather than on the quality of education, despite the increasing availability of cross-national assessment data on the continent. The two primary sources of cross-national data are the Southern and Eastern African Consortium for Monitoring Educational Quality (SACMEQ) and the Programme d’Analyse des Systèmes Éducatifs de la CONFEMEN (Programme for the Analysis of Education Systems of CONFEMEN Countries; PASEC). (CONFEMEN denotes the Conference of Education Ministers of French-Speaking Countries.) The former assesses countries in Southern and East Africa, while the latter assesses countries in francophone West Africa and a few other countries. A disconnected discourse, focusing on one issue or the other, is problematic for two reasons. First, observing the quantity of education without regard for its quality clouds the analysis, primarily because the underlying assumption that enrolment and attainment are correlated with learning is often not true, as this chapter will demonstrate. Second, analysing educational outcomes for those attending school without considering the enrolment and dropout profiles of the countries under review is likely to bias the results. Developing countries with lower enrolments and higher dropout rates perform better, on average, than otherwise similar countries that have higher enrolments and fewer dropouts (UNESCO, 2005). This phenomenon is largely due to the selection effects involved: The ‘strongest’ (that is, the wealthiest, 49

SPAULL & LILENSTEIN

most advantaged, and most able) students enrol and then remain in the schooling system (Lambin, 1995). The only way to correct for these biases is to find some method of combining measures of both access and quality into an integrated statistic. This was precisely the aim of our earlier work in Southern Africa (Spaull and Taylor, 2015). In that study, we developed a method – which we have also applied in this chapter – for creating a composite statistic of educational quantity and educational quality by combining household data (Demographic and Health Survey; DHS) on grade completion and survey data (SACMEQ) on cognitive outcomes for 11 African countries. The indicator that we created was called ‘access to literacy’ and measured the proportion of children from a particular cohort that both completed Grade 6 and acquired basic literacy skills. The present study extends this work by shifting the focus to two West African countries – Burkina Faso and Senegal – and using PASEC data instead of SACMEQ data. DATA AND METHOD

In this chapter, we use the latest publicly available data from PASEC as well as matched-cohort DHS data. PASEC differs from SACMEQ in that PASEC tests both Grade 2 students and Grade 5 students at the beginning and end of the year, while SACMEQ tests only Grade 6 students and only at the end of the year. In this chapter, we use end-of-year assessments only. PASEC has developed three learning benchmarks based on the percentage of items answered correctly rather than any psychometric properties or aggregation (as in SACMEQ). The PASEC assessment has a multiple choice format; one can score 30% simply by answering the questions randomly. Based on this fact, PASEC and CONFEMEN separate achievement on the assessment into levels. PASEC Level 1 is a score between 0% and 24% correct answers. This score is lower than what could be obtained by answering randomly; students scoring at this level are considered to be failing scholastically. Level 2 is a score between 25% and 40%. This level encompasses students scoring just below and just above what a series of random answers would give. Thus, these students are demonstrating basic knowledge barely, if at all. At this level, according to PASEC and CONFEMEN, students are not failing, but they also do not possess basic knowledge of reading and writing. Level 3 is the final level and includes all scores above 40%. At this level, students are said to possess basic knowledge of reading and writing (Education Policy and Data Center, 2012). The threshold of 40% is barely above a score that can be achieved by guessing. This is a low threshold of basic literacy and statistics gained using this threshold can be viewed as an upper-bound estimate (i.e., we are confident that we are not underestimating literacy levels but not confident that we are not overestimating them). It has also been used before with earlier rounds of PASEC surveys (Michaelowa, 2001). The PASEC tests are broadly grade-aligned; a score that falls within the Level 3 benchmark indicates that the student possesses basic knowledge of reading and 50

ASSESSING EARLY LITERACY OUTCOMES

writing that is suitable for that grade level. Hence, we consider a Grade 2 child who has reached Level 3 to have acquired ‘emergent literacy’; a Grade 5 child who has reached Level 3 is seen as having acquired ‘basic literacy’. These distinctions are somewhat arbitrary given PASEC’s own arbitrary assignment to levels. However, since a Grade 2 student is just beginning to read and a Grade 5 student should be reading at this level, we think this nomenclature is appropriate. In Burkina Faso, PASEC 2006 tested 2,116 Grade 2 students and 2,221 Grade 5 students from 158 primary schools drawn from a nationally representative sample. In Senegal, PASEC 2006 tested 1,979 Grade 2 students and 1,910 Grade 5 students from 180 primary schools also drawn from a nationally representative sample. To find corresponding household survey data, in order to assess grade completion, we used the DHS data from 2010 in Burkina Faso and from 2012–13 in Senegal. In Burkina Faso, 14,424 households were surveyed, while in Senegal 4,175 households were surveyed. In most African countries, the practices of delayed enrolment and grade repetition are widespread, leading to a considerable proportion of older students in earlier grades. For example, in Burkina Faso, 10% of those currently enrolled in Grade 5 in DHS 2010 were age 14 or above, despite the fact that students in this grade should be 10 years old. Thus, in order to find an appropriate age cohort, one needs to select those ages where everyone who might be expected to complete Grade 5 has already done so. For example, one cannot use an age cohort such as 15- to 17-year-olds since many 15- and 16-year-olds may go on to complete Grade 5, even though they have not completed it yet. Using a cohort aged 15 to 17 will thus underestimate the true Grade 5 completion rate. With this in mind, we chose cohorts for this study by selecting the youngest age cohort (with a five-year grouping) in which at least 95% of the cohort was not currently enrolled in Grade 5 (or in Grade 2 for the Grade 2 cohort). For Grade 5, the cohort is 16 to 20 years in both Burkina Faso and Senegal. For Grade 2, the cohorts are 11 to 15 years for Burkina Faso and 12 to 16 years for Senegal. DHS datasets were chosen to match these age cohorts. For example, the average Burkinabe is 12 years old in Grade 5 and the PASEC data were collected in 2006; those students would be 16 years old in 2010 and 20 years old in 2014. Hence, a DHS dataset collected in 2010 matches our chosen age cohort relatively well. The DHS dataset that best matches the Grade 2 Burkinabe cohort was also the 2010 DHS dataset. To match Senegal Grade 2 and Grade 5 we used the 2012–2013 DHS dataset. For a more detailed discussion on cohort matching using these two countries, see Lilenstein (2016). Overall, the method applied here is not technically complex and in many instances simply involves multiplying the percentage of a cohort that completes Grade 5 with the percentage of the PASEC sample that acquires basic literacy skills in Grade 5. For example, our data show that 60% of a cohort of Senegalese children completes Grade 5 (DHS), but only 37% of Grade 5 students acquire basic literacy skills (PASEC). Thus, the access-to-basic-literacy rate would be 22% (60% x 37% = 22%): Only 22% 51

SPAULL & LILENSTEIN

of a cohort of Senegalese children will complete Grade 5 and acquire basic literacy skills. Notice how both the grade completion rate (60%) and the PASEC literacy rate (37%) both overestimate the success of the schooling system when they are seen in isolation. We argue that the access-to-literacy rate is the best indicator of school system performance, since it accounts for both access (quantity) and learning (quality). Because of space constraints, we will not repeat the technical details explaining how the method accounts for differential enrolment (and subsequent sampleselection) in the PASEC survey. This involves identifying the proportion of the DHS sample that makes it into the PASEC sampling frame, given that this differs by socio-economic status, and its interaction with gender. We direct the reader to Spaull and Taylor (2015) for a detailed explanation. RESULTS

The national level results that combine measures of both grade completion (from DHS) and learning (from PASEC) appear in Figure 1 for the Grade 2 cohort, and Figure 2 for the Grade 5 cohort. From these graphs, we see that the levels of nonenrolment are high in both countries, but especially in Burkina Faso. The Grade 2 cohort in Burkina Faso sees only two in every ten (23%) learners achieving basic literacy. Senegal’s cohort only does slightly better at three in every ten (34%).

Figure 1. National levels of access and quality of education – Grade 2 literacy

52

ASSESSING EARLY LITERACY OUTCOMES

Looking at the Grade 5 cohort in Burkina Faso, we see that, of 10 Burkinabes, five will not enrol in school, one will enrol but drop out before Grade 5, three will complete Grade 5 but not acquire basic literacy skills, and only one (11%) will complete Grade 5 and acquire basic literacy skills.

Figure 2. National levels of access and quality of education – Grade 5 literacy

In both cohorts and in both countries, between 25% and 34% of children do complete the respective grade (either Grade 2 or 5) but do not acquire either emergent literacy skills (Grade 2) or basic literacy skills (Grade 5). This points to the absence of learning even among those who are enrolled and do progress into higher grades. It is also worth noting that, in Burkina Faso, fewer than 25% of a cohort acquire gradeappropriate skills in either Grade 2 or Grade 5; in Senegal the percentage is slightly higher only among the Grade 2 cohort, at 34%. Access: Quantity of Education Table 1, focusing on Grade 2 data, and Table 2, focusing on Grade 5 data, summarize the rates for access, literacy, and access-to-literacy for Burkina Faso and Senegal. The first and second columns of Tables 1 and 2 show non-enrolment and dropout rates among students in Burkina Faso and Senegal for the Grade 2 and 5 cohorts respectively. In comparing the rates of those who never enrol with those who drop out, it becomes clear that, as one might expect, there is almost no dropping out at the Grade 2 level;

53

SPAULL & LILENSTEIN

even at the Grade 5 level, the levels of dropout pale in comparison to the incredibly high levels of non-enrolment. With both the Grade 2 cohort and the Grade 5 cohort in Burkina Faso, 59–76% of the poorest group of students will never enrol in school. By comparison, only 15–25% of the richest group of Burkinabes or Senegalese will not enrol in school. Non-enrolment among the poorest group in Senegal is lower than it is in Burkina Faso, with 43–53% of the poorest group never enrolling in school. While many studies have looked at the large disadvantages that girls face relative to boys in some countries (Lewis and Lockheed, 2006), the evidence in Tables 1 and 2 shows that these are minuscule in comparison to the inequalities between the wealthiest 20% of children and the poorest 40% of children. In Burkina Faso, the gap in Grade 2 completion between boys (57%) and girls (52%) is small (completion rates are shown in the third column of Tables 1 and 2). In stark contrast, the wealthiest group of Burkinabes is twice as likely to complete Grade 2 (82%) as the poorest group (38%). The gender gap widens in Grade 5, but is still small relative to the wealth gap, which increases to four times. The situation is better in Senegal, but the wealthiest group of students is still more than one and a half times as likely to complete Grade 5 (71%) as the poorest group (40%). Looking at the ‘double disadvantage’ (Lewis and Lockheed, 2006) faced by girls who are also poor – and who are therefore part of two groups that are individually and jointly disadvantaged – we compared the poorest 40% of girls with the poorest 40% of boys. Although there is some disadvantage to being poor and female, this ranges between 6 and 7 percentage points in Burkina Faso and from 0 to 5 percentage points in Senegal. In Burkina Faso, this could represent a floor effect: access rates are so low among the poor (16% in Grade 5) that there is little room for further inequalities to emerge between boys and girls. It is interesting that although neither Burkina Faso nor Senegal displays large gender inequalities overall or among the poor, gender inequality is evident among the rich in both countries. In Burkina Faso, only 59% of wealthier females will complete Grade 5, compared with 77% of wealthier males; in Senegal, 62% of wealthier females will complete Grade 5, compared with 81% of wealthier males. This gap is smaller among the Grade 2 cohorts, ranging between 5 and 10 percentage points. This could be because the cohort is a more recent cohort, or because the gap only emerges in higher grades. Learning: Quality of Education If one limits the analysis to the PASEC sample and asks simply how much learning is taking place among those students who do complete Grade 2 or Grade 5, the answer is not encouraging. These results appear in the fourth column of Tables 1 and 2. Fewer than half of all students assessed in PASEC in these two countries reached Level 3 (emergent literacy in Grade 2 and basic literacy in Grade 5); the exceptions are the richest group in Senegal, and for Grade 2, females in Senegal and the richest females in Burkina Faso. The gaps in learning outcomes between the richest and 54

2. Drop out before Gr2

3. Completed Gr2

4. PASEC emergent literacy rates (Gr2)

5. Completed Gr2 but did not reach emergent literacy

6. Completed Gr2 with emergent literacy

41 (1.6)

10 (1.4)

21 (1.6)

Rich20M

Rich20F

17 (3.8)

13 (3.2)

24 (2.6)

24 (2.6)

43 (5.0)

44 (2.9)

15 (2.9)

24 (2.2)

43 (3.5)

30 (2.2)

30 (1.9)

30 (1.7)

1 (0.5)

2 (0.5)

3 (0.4)

3 (0.4)

3 (0.4)

3 (0.4)

2 (0.3)

3 (0.3)

3 (0.3)

3 (0.3)

3 (0.3)

3 (0.2)

0 (0.1)

2 (1.6)

2 (0.5)

1 (0.5)

3 (0.6)

3 (0.7)

1 (0.8)

1 (0.3)

3 (0.4)

2 (0.3)

2 (0.4)

2 (0.3)

78 (1.6)

88 (1.5)

56 (1.6)

61 (1.6)

35 (1.5)

41 (1.5)

82 (1.2)

58 (1.3)

38 (1.2)

52 (1.1)

57 (1.2)

55 (1.0)

82 (3.9)

87 (4.2)

78 (2.6)

77 (2.6)

54 (4.5)

54 (2.9)

84 (3.2)

78 (2.2)

54 (3.3)

68 (2.2)

67 (1.9)

68 (1.8)

53 (4.2)

49 (3.8)

37 (5.4)

43 (4.5)

32 (5.1)

37 (4.3)

51 (3.5)

40 (4.5)

35 (4.0)

40 (3.3)

43 (2.9)

42 (2.9)

70 (4.7)

74 (4.0)

49 (3.5)

46 (3.9)

39 (6.7)

33 (7.1)

72 (3.7)

48 (3.0)

36 (5.8)

51 (3.3)

49 (3.2)

50 (2.9)

37 (4.5)

45 (4.1)

35 (5.7)

35 (4.8)

24 (5.3)

26 (4.6)

40 (3.7)

35 (4.7)

25 (4.2)

31 (3.5)

32 (3.2)

32 (3.0)

25 (6.1)

23 (5.8)

42 (4.7)

43 (4.7)

33 (8.1)

36 (7.6)

23 (4.9)

41 (3.7)

34 (6.7)

33 (3.9)

34 (3.8)

34 (3.4)

41 (4.5)

43 (4.1)

21 (5.7)

26 (4.8)

11 (5.3)

15 (4.6)

42 (3.7)

23 (4.7)

13 (4.2)

21 (3.5)

25 (3.2)

23 (3.0)

57 (6.1)

64 (5.8)

36 (4.7)

34 (4.7)

21 (8.1)

18 (7.6)

61 (4.9)

37 (3.7)

20 (6.7)

35 (3.9)

33 (3.8)

34 (3.4)

Notes: 1. Values shown are percentages. 2. Standard errors are given in parenthesis. 3. Column 6 represents the access-to-basic-literacy rate.

36 (1.6)

Mid40F

16 (1.2)

Rich20

Mid40M

39 (1.3)

Mid40

56 (1.5)

59 (1.3)

Poor40

63 (1.5)

45 (1.1)

Females

Poor40F

40 (1.1)

Males

Poor40M

43 (1.0)

Burkina F. Senegal Burkina F. Senegal Burkina F. Senegal Burkina F. Senegal Burkina F. Senegal Burkina F. Senegal

National

Sub group

1. Never enrolled

Table 1. Access, literacy, and access-to-literacy rates for Burkina Faso and Senegal in Grade 2

ASSESSING EARLY LITERACY OUTCOMES

55

56

71 (1.8)

81 (1.4)

49 (2.0)

68 (1.7)

14 (1.5)

33 (1.7)

Mid40M

Mid40F

Rich20M

Rich20F

8 (1.0)

9 (1.3)

8 (0.8)

12 (1.0)

7 (0.8)

9 (1.0)

8 (2.5)

4 (1.3)

9 (1.2)

9 (1.5)

9 (1.2)

9 (1.0)

59 (1.9)

77 (2.0)

24 (1.5)

40 (1.8)

13 (1.2)

20 (1.4)

67 (1.4)

32 (1.3)

62 (5.0)

81 (4.1)

61 (2.7)

65 (2.5)

38 (3.5)

43 (3.1)

71 (3.5)

63 (2.1)

38 (3.6)

44 (4.0)

22 (3.6)

29 (3.4)

23 (3.5)

25 (3.9)

41 (3.5)

26 (2.9)

62 (5.6)

58 (4.6)

35 (3.4)

37 (3.1)

17 (4.9)

21 (4.5)

58 (4.0)

38 (3.2)

37 (4.1)

43 (4.5)

19 (3.9)

29 (3.8)

10 (3.7)

15 (4.2)

39 (3.8)

24 (3.2)

12 (3.2)

22 (2.9)

24 (7.5)

34 (6.1)

40 (4.3)

41 (4.0)

31 (6.0)

34 (5.4)

30 (5.4)

39 (3.8)

33 (5.0)

33 (3.8)

22 (4.1)

34 (4.5)

5 (3.9)

11 (3.8)

3 (3.7)

5 (4.2)

28 (3.8)

8 (3.2)

4 (3.2)

9 (2.9)

38 (7.5)

47 (6.1)

21 (4.3)

24 (4.0)

7 (6.0)

9 (5.4)

41 (5.4)

24 (3.8)

7 (5.0)

19 (3.8)

Notes: 1. Values shown are percentages. 2. Standard errors are given in parenthesis. 3. Column 6 represents the access-to-basic-literacy rate.

30 (4.5)

18 (3.2)

32 (3.1)

27 (2.6)

54 (3.9)

51 (3.3)

6 (1.4)

9 (0.9)

18 (4.3)

36 (3.1)

22 (3.6)

20 (3.2)

Poor40F

9 (0.8)

10 (0.7)

24 (3.0)

29 (2.7)

14 (3.0)

12 (2.6)

Poor40M

24 (3.3)

30 (2.5)

40 (2.6)

52 (2.1)

36 (3.6)

35 (3.2)

25 (1.3)

16 (1.1)

31 (1.1)

28 (3.0)

24 (2.6)

58 (1.5)

9 (0.8)

9 (0.9)

38 (2.8)

37 (2.6)

Rich20

8 (0.7)

8 (0.5)

34 (2.7)

32 (2.4)

Mid40

53 (2.8)

40 (2.2)

58 (2.2)

55 (1.7)

76 (1.4)

42 (1.3)

36 (1.0)

62 (1.2)

8 (0.8)

8 (0.6)

Poor40

10 (0.7)

9 (0.4)

Females

34 (2.1)

37 (1.8)

48 (1.4)

6. Completed Gr5 with basic literacy

Males

5. Completed Gr5 but did not reach basic literacy

55 (1.1)

4. PASEC basic literacy rates (Gr5)

Burkina F. Senegal Burkina F. Senegal Burkina F. Senegal Burkina F. Senegal Burkina F. Senegal Burkina F. Senegal

3. Completed Gr5

National

2. Drop out before Gr5

Sub group

1. Never enrolled

Table 2. Access, literacy, and access-to-literacy rates for Burkina Faso and Senegal in Grade 5

SPAULL & LILENSTEIN

ASSESSING EARLY LITERACY OUTCOMES

poorest groups are not as large in Burkina Faso as they are in Senegal. Among the richest group in the Grade 5 cohort in Burkina Faso, 41% acquired basic literacy compared to 24% in the poorest group. In Senegal, the gap is larger: the percentage of Grade 5 students acquiring basic literacy in the richest group (58%) is three times the percentage in the poorest group (18%). Figure 3 shows the breakdown of grade completion and learning outcomes for the various subgroups in Senegal for the Grade 5 cohort (16- to 20-year-olds). Here, the marked difference in learning outcomes between the wealthiest 20% of children and the remaining 80% is apparent. The wealthiest group of children have very low non-enrolment rates and even lower dropout rates. While approximately 35% of children complete Grade 5 but still do not acquire basic reading skills, the percentage of children who do complete Grade 5 and acquire basic literacy skills among the richest group (41%) is twice as high as the national average (20%) and five times as high as among the poorest group (7%). A similar pattern is evident in Burkina Faso (Figure 4). All the groups perform worse than their counterparts in Senegal. It is interesting to note that although the percentage of the wealthier group that completes Grade 5 and acquires basic literacy skills is low (28%), it is still seven times the corresponding percentage for the poorest group, which is only 4%.

Figure 3. Grade 5 completion and learning outcomes by sub-groups for the cohort of 16- to 20-year-olds in Senegal (Grade 5 cohort)

57

SPAULL & LILENSTEIN

Figure 4. Grade 5 completion and learning outcomes by sub-groups for the cohort of 16- to 20-year-olds in Burkina Faso (Grade 5 cohort)

Access to Learning: Combining Quantity and Quality of Education The low levels of learning and the high levels of non-enrolment are sobering in and of themselves; when combined, they depict a crisis in education in these two countries. In Burkina Faso, the access-to-emergent-literacy rate (in the last columns of Tables 1 and 2) is only 23%: only one in four children will complete Grade 2 and acquire emergent literacy skills. The access-to-emergent-literacy rate in Senegal is only 34%: one in three children will complete Grade 2 and acquire emergent literacy skills. Looking at the slightly older cohort, the results are even more concerning. The access-to-basic-literacy rate in Burkina Faso is only 12%; that is, only 12% of a cohort will complete Grade 5 and acquire basic literacy skills. In Senegal, only 20% of a cohort will complete Grade 5 and acquire basic literacy. Among the richest group of students, the access-to-emergent-literacy rate in Burkina Faso (42%) is more than double the national average. In Senegal, the accessto-emergent-literacy among the richest group (61%) is also double the national average. The inequalities are larger at the higher grades. As mentioned previously, the access-to-basic-literacy rate in Senegal is five times as high for the richest group (41%) as for the poorest group (7%). In other words, two in five students in the wealthy group will complete Grade 5 and acquire basic literacy skills, compared to less than one in 10 in the poorest group. In Burkina Faso, only one in 20 children (4%) from the poorest 40% of the population completes Grade 5 and acquires basic literacy skills. It is a real possibility that not a single student among the poorest 58

ASSESSING EARLY LITERACY OUTCOMES

40% of Burkinabe children completed Grade 5 and acquired basic literacy skills. This is because the standard error corresponding to the 4% estimate is 3.2; the 95% confidence interval therefore overlaps with zero. In other words, once we take sampling variation into account, the confidence band for this estimate ranges from 0 to 10%. In our earlier research on sub-Saharan Africa (Spaull and Taylor, 2015; Taylor and Spaull, 2015), we included both Malawi and Mozambique. Both countries have low levels of economic and social development and may appear comparable to Burkina Faso or Senegal. For example, of the 188 countries included in the United Nations Development Program’s Human Development Index, all four of these countries rank in the lowest 5% of the index (UNDP, 2015). Yet, even in Mozambique – the SACMEQ country with the lowest access-to-literacy rate – 17% of children among the poorest group completed Grade 6 and acquired basic literacy skills. In Malawi, the corresponding figure is 25%. These figures are much higher than the 6% for the poorest group in Burkina Faso or the 8% for Senegal. Having said this however, unfortunately the SACMEQ and PASEC levels are not currently comparable. One might expect the SACMEQ basic literacy level to be more difficult, given that it is assessing a higher grade and that the levels are psychometrically calibrated, but this is speculation. Before we reach the conclusion section, two technical issues are worth highlighting. First, there are currently no universally shared definitions of ‘literacy’, ‘emergent literacy’, or ‘basic literacy’. As a result, estimates here cannot be compared to those using SACMEQ or any other data. This problem, which is ongoing, needs to be resolved before any meaningful global learning goals can be measured or monitored. To that end, we encourage organizations such as CONFEMEN, SACMEQ, UNESCO, and similar institutions, to collaborate with the goal of sharing and linking literacy test items and definitions of literacy. Second, we intend to conduct this same analysis using more recent PASEC data (for example, 2014 data). Given the extremely low levels of basic reading proficiency, the hope is that considerable improvements have occurred between PASEC 2006 and 2014. Nevertheless, this research provides a useful baseline against which to compare any progress, whether in access, learning or equity between socioeconomic groups. CONCLUSION

The picture that emerges from this analysis, based on the DHS and PASEC data, is truly dire. In the most recent cohort that we assessed in Burkina Faso, more than a third had never been to school, another third completed Grade 2 without even the most elementary emergent literacy skills, and only one in four children (23%) had completed Grade 2 with emergent literacy. The figures for Senegal are only marginally better, with 30% of the cohort reporting that they had never enrolled in school, while 34% completed Grade 2 without acquiring emergent literacy skills, 59

SPAULL & LILENSTEIN

and only one in three children (34%) completing Grade 2 with emergent literacy skills. Since the Grade 5 cohort that we examined were 16 to 20 years old (in 2010 for Burkina Faso and in 2012–13 for Senegal), and since access to education has expanded considerably in these countries (especially Burkina Faso), the figures are even lower for this cohort. Nevertheless, it is truly worrying that among the 2010 cohort of Burkinabes 16 to 20 years old, half had never been to school, 9% dropped out before Grade 5, 25% completed Grade 5 without basic literacy skills, and only 11% completed Grade 5 with basic literacy skills. Among the Senegalese cohort, only 21% completed Grade 5 with basic literacy skills. With some fairly conservative assumptions, one can say that at least 75% of these older cohorts did not acquire – and, arguably, will not acquire – basic reading and writing skills. It is difficult to appreciate how this can be in the 21st century. The only thing more troubling than the national averages is the socio-economic inequalities that they hide. Among the poor in both countries, at most 20% of the youngest cohort will complete Grade 2 with emergent literacy and fewer than 10% of the older cohort will complete Grade 5 with basic literacy skills. The wealthiest 20% of Burkinabe and Senegalese children are six to seven times as likely to complete Grade 5 with basic literacy skills as the poorest 40% in each country. While the access-to-emergent-literacy rate among the poorest group of Burkinabe girls is 13%, the corresponding figure for the wealthiest Burkinabe girls is 42%. Among Senegalese girls, the comparable figures are 20% and 61% respectively. The analysis presented is particularly pertinent to discussions around the UN Sustainable Development Goals (SDGs). On one hand, this analysis highlights the importance of such goals, which prioritize both access to education and, as well, learning outcomes. On the other hand, they call into question the validity of a single undifferentiated goal, indiscriminately applied to Burkina Faso and Brazil, Senegal, and South Africa. We fully agree that countries such as the two that are the focus of this chapter require prioritized support and funding from the international community, and even that the rate of progress here must be much more rapid than in other countries. At the same time, we caution that the only thing worse than setting an unambitious goal is setting one so unrealistic that no one believes that it can or will be achieved. Setting an unrealistic goal provides absolution from any responsibility or accountability: no one believes that the goal can or will be achieved, and consequently no one is held responsible when it is not achieved. Setting ambitious but plausibly achievable goals is the most prudent road ahead. With sustained political and financial support for the early years of schooling, and specifically for improvements in the quality of teaching early reading, it is not unreasonable to think that rapid progress is possible. Ensuring that all children learn to read and write in the early years of primary school is necessary, not only for their educational development, but also for the economic, social, and cultural development of the countries in which they live.

60

ASSESSING EARLY LITERACY OUTCOMES

REFERENCES African Commission on Human and Peoples’ Rights (1981). African charter on human and peoples’ rights. http://www.achpr.org/files/instruments/achpr/banjul_charter.pdf Barro, R. J., & Lee, J. (2013). A new data set of educational attainment in the world, 1950–2010. Journal of Development Economics, 104, 184–98. Basu, A. M. (2002). Why does education lead to lower fertility? A critical review of some of the possibilities. World Development, 30, 1779–1790. Currie, J. (2009). Healthy, wealthy, and wise: Socioeconomic status, poor health in childhood, and human capital development. Journal of Economic Literature, 47, 87–122. Dembélé, M., Somé, T. H., & Ouédraogo, F. (2015a). Burkina Faso: An overview. In E. J. Takyi-Amoako (Ed.), Education in West Africa. New York, NY: Bloomsbury. Dembélé, M., Somé, T. H., & Ouédraogo, F. (2015b). Burkina Faso: Trends and futures. In E. J. TakyiAmoako (Ed.), Education in West Africa. New York, NY: Bloomsbury Academic. Easterlin, R. A. (1981). Why isn’t the whole world developed? Journal of Economic History, 41, 1–19. Education Policy and Data Center (2012). SACMEQ and PASEC. http://www.epdc.org/data-about-epdcdata-epdc-learning-outcomes-data/sacmeq-and-pasec Filmer, D., Hasan, A., & Pritchett, L. (2006). A millennium learning goal: Measuring real progress in education. Working paper 97. Washington, DC: Center for Global Development. Glewwe, P. (2002). Schools and skills in developing countries: Education policies and socioeconomic outcomes. Journal of Economic Literature, 40, 436–82. Goldin, C. D., & Katz, L. F. (2009). The race between education and technology. Cambridge, MA: Harvard University Press. Hanushek, E. A., & Woessmann, L. (2008). The role of cognitive skills in economic development. Journal of Economic Literature, 46, 607–68. Heyneman, S. P. (2003). Education, social cohesion, and the future role of international organizations. Peabody Journal of Education, 78, 25–38. Lambin, R. (1995). What can planners expect from international quantitative studies? In B. Wilfried & R. H. Lehmann (Eds.), Reflections on educational achievement: Papers in honor of T. Neville Postlethwaite. New York, NY: Waxmann. Lewis, M., & Lockheed, M. (2006). Inexcusable absence. Washington, DC: Center for Global Development. Lilenstein, A. (2016). Integrating indicators of education quantity and quality in six francophone African countries (Unpublished master’s thesis). University of Cape Town, Cape Town. Majgaard, K., & Mingat, A. (2012). Education in sub-Saharan Africa: A comparative analysis. Washington, DC: World Bank. Manion, C. (2015). Senegal: Trends and futures. In E. J. Takyi-Amoako (Ed.), Education in West Africa. New York, NY: Bloomsbury Academic. Michaelowa, K. (2001). Primary education quality in francophone sub-Saharan Africa: Determinants of learning achievement and efficiency considerations. World Development, 29, 1–34. Nussbaum, M. C. (2006). Education and democratic citizenship: Capabilities and quality education. Journal of Human Development, 7, 385–95. Pritchett, L. (2013). The rebirth of education: Schooling ain’t learning. Washington, DC: Center for Global Development. Pritchett, L., & Beatty, A. (2015). Slow down, you’re going too fast: Matching curricula to student skill levels. International Journal of Educational Development, 40, 276–288. Salmi, J. (2000). Violence, democracy and education: An analytical framework. LCSHD Paper Series. Washington, DC: World Bank Salmon, L., & Dramani, L. (2015). Senegal: An overview. In E. J. Takyi-Amoako (Ed.), Education in West Africa. New York, NY: Bloomsbury Academic. Sen, A. (1999). Development as freedom. Oxford: Oxford University Press. Spaull, N., & Taylor, S. (2015). Access to what? Creating a composite measure of educational quantity and educational quality for 11 African countries. Comparative Education Review, 59, 133–165.

61

SPAULL & LILENSTEIN Taylor, S., & Spaull, N. (2015). Measuring access to learning over a period of access to schooling: The case of southern and eastern Africa since 2000. International Journal of Educational Development, 41, 47–59. UNESCO (2005). Education for All: The quality imperative. EFA global monitoring report. Paris: UNESCO. United Nations (1948). The universal declaration of human rights. http://www.ohchr.org/EN/UDHR/ Documents/UDHR_Translations/eng.pdf UNDP [United Nations Development Programme] (2015). Human development report 2015: Work for human development. New York, NY: UNDP.

62

ELIZABETH J. PRETORIUS

4. GETTING IT RIGHT FROM THE START Some Cautionary Notes for Early Reading Instruction in African Languages

INTRODUCTION

High levels of literacy are needed to do well in school and to survive and thrive in the high-tech digital world of the 21st century. Yet, large numbers of children in lowincome schools in developing countries continue to have low literacy levels, putting them at risk of never getting a fair start in life, and never breaking out of the cycle of poverty. Large-scale national and international assessments, such as Progress in International Reading Literacy (PIRLS), Trends in International Mathematics and Science Study (TIMSS), and the Southern and Eastern African Consortium for the Monitoring of Educational Quality (SACMEQ), consistently reveal disparities in literacy and numeracy accomplishments in children from disadvantaged communities in developing countries. Poor reading leads to mediocre school achievement at best or, at worst, downright school failure and eventual dropout. Poverty is a risk factor in literacy development, as is learning to read in a language unfamiliar to a child. Other macrosystemic factors associated with low literacy include poor governance in schools, poorly resourced schools, and inadequately trained teachers. What other factors can also pose a risk for literacy development, specifically reading literacy? To date, little research has been done on reading in African languages. Because of its historical ties with Europe, and because English, French, and Portuguese are the main colonial languages, Africa has tended to look to reading research from the North and West. These views of reading are often adopted unwittingly in developing countries, producing what may be inappropriate reading applications in unsuitable contexts. For example, instructional methods for teaching early reading in English may be adopted for early reading in African languages without consideration of how the reading principles underlying the English-based reading methods could best be adapted to the African language in question. Launching children on successful reading trajectories from the start of schooling is a priority. A faltering initial reading trajectory creates cracks in literacy development which ‘in time become gaps, and finally … chasms in learning’ (Johnson, 2012). Inappropriate applications of reading instruction across diverse linguistic contexts may pose risks to early reading development.

© JOINTLY BY IBE-UNESCO AND KONINKLIJKE BRILL NV, LEIDEN, 2019 DOI: 10.1163/9789004402379_005

PRETORIUS

In this chapter, I argue that the use of appropriate reading methods can help to launch readers on successful reading trajectories from the start of their schooling, provided that we have a principled basis for determining what is appropriate in the reading context. I propose that the unique linguistic and orthographic nature of the languages in which reading occurs should provide the basis for guidelines for early instruction, specifically when the focus is on developing early decoding skills (that is, the development of fluent code-based skills on which meaningful comprehension relies). A FRAMEWORK FOR GUIDING READING INSTRUCTION

Far more is now known about reading than was known a few decades ago. We now know more about the cognitive-linguistic complexities of reading, its neurological underpinnings, the close association between reading and affective factors, and the socially constituted nature of literacy practices and values. We also have a fairly good idea of what is important in early literacy instruction, especially in alphabetic languages such as English and other Western languages. In the United States, the National Reading Panel (2000) has identified the five ‘pillars’ of reading as essential components that increase the chances of success in reading, at least in alphabetic writing systems. These five ‘pillars’ are phonemic awareness, phonics, fluency, vocabulary, and comprehension. The body of knowledge that has accrued from 70 years of research is considerable. However, it would be naïve to assume that there is consensus in the field or that we fully understand all the variables that have an impact on literacy development. Literacy remains a contested and politicized domain; worldwide, education systems face persistent challenges in providing quality literacy development for atrisk children from disadvantaged communities. How then are we to find our way to address the unequal literacy trajectories that are characteristic of high poverty contexts? First, we need to be mindful of the sources of our knowledge about reading. From which languages do research findings derive? Second, given the complexity of reading, we need to be reasonably sure that there is converging evidence for the claims that are made about reading. Third, we need to be mindful of how we apply what we know and understand about reading in contexts that are different from the original contexts. Will the implied predictions or outcomes still hold true in a different context? Much of what we know and understand about reading comes from a large body of research based on reading in English and other European languages, among them, in particular, Dutch, Spanish, Italian, French, German, Greek, and Swedish. These are languages that belong to the Germanic or Romance language families. They are classified typologically on a continuum, according to whether they have greater or fewer analytic or inflectional morphosyntactic features (that is, whether they convey meaning primarily through syntax, especially word order, or by adding 64

EARLY READING INSTRUCTION IN AFRICAN LANGUAGES

prefixes and suffixes to words). Increasingly, research into reading in languages that are typologically different, such as Turkish, Finnish, and Basque, is entering academic reading journals. These languages differ from European languages in that they are agglutinating languages. In these languages, words consist of sequences of morphemes attached to a stem, each of which represent a different grammatical category, as in the word abayifundi ‘they aren’t reading it’, which consists of five morphemes a+ba+yi+fund(stem)+i (explained in greater detail in the following section). Is reading in an agglutinating language the same as, or different from, reading in analytic or inflectional languages? Furthermore, the study of reading research on languages that do not have an alphabetic writing system is increasing; Chinese and Japanese and other Asian or Indian languages are examples of such languages. Research suggests that different writing systems can affect the reading process in different ways (Wolf, 2008). Although the languages of Western Europe all have an alphabetic orthography (that is, writing system) in common, they differ as to how transparent or opaque their orthographies are. In transparent orthographies, the relationship between letters and their sounds are regular (e.g., the letter u always stands for the sound /u/ in Xhosa), whereas in opaque orthographies, the letter-sound relationship is less predictable (e.g., in English the letters -ou- represent different vowel sounds in though R‫ݜ‬ through /u/, rough ‫ ݞ‬bought ‫ܧ‬, hiccough ԥ, plough D‫ ݜ‬ Besides typological and orthographic differences, reading research from the North and West also tends to address reading in fairly affluent, print-rich societies. These societies have relatively high levels of literacy in the adult population in general, and fairly long traditions of literacy and mass education. In contrast, teachers in developing countries often have low levels of general education (O’Sullivan, 2002). They have large classes and few print-based resources, and they view literacy in functional terms and tend to be reluctant readers themselves, reading mainly for work purposes (Pretorius and Knoetze, 2012; Taylor and Taylor, 2013). Furthermore, the SACMEQ results indicate that teachers often do not perform well on the same literacy and numeracy tests that their learners take (Spaull, 2013), and assumptions may differ as to what good teaching and learning entail. Further international reading research is still needed to identify critical school and classroom features and effective practices that promote sound literacy development across different sociocultural and socio-economic contexts. In order to avoid applying knowledge of reading inappropriately in contexts that are different, I suggest in this chapter a framework for reading that includes considering the nature of the alphabetic writing system (e.g., transparent/opaque and conjunctive/disjunctive orthographies) and the typology of the language (e.g., agglutinating languages). By no means does a focus on the linguistic and orthographic features of language imply that this is the most important aspect underlying early reading success. Indeed, many individual, social, socio-economic, and pedagogic factors converge in developing literacy skills in children. However, a discussion of these other factors is beyond the scope of this chapter. 65

PRETORIUS

IDENTIFYING WHAT IS DIFFERENT: STRUCTURAL AND ORTHOGRAPHIC FEATURES OF LANGUAGE

In linguistic typology, languages can fall into two broad categories, isolating/analytic and inflectional/synthetic languages. Isolating/analytic languages convey meaning largely through grammar and word order, while inflectional/synthetic languages convey meaning through the use of morphemes. A morpheme is the smallest meaningful linguistic unit; an example is the suffix -ed in English that conveys pastness. English, for instance, is regarded as a mildly inflectional language. In its development from Old English to Modern English, it has become increasingly analytic, shedding many of its inflectional features; only a few inflected forms remain (for example, the past tense marker -ed as in gallop-galloped, and the plural marker -s as in horse-horses). At the other end of the inflectional continuum and part of the inflectional/synthetic group are the agglutinating languages, which are heavily inflectional (that is, they have many morphemes). These include Finnish, Turkish, and Basque, as well as African languages south of the equator, which belong to a large language family known as the Southern Bantu languages. Agglutinating languages are morphologically rich languages that rely heavily on affixation: prefixes, infixes, and suffixes are added to noun and verb stems in sequence to convey grammatical and semantic information. In the following sections, the focus will be mainly on the linguistic and orthographic differences between English, as a partially analytic language, and Southern African languages such as Xhosa and Zulu as agglutinating languages, and why these differences are important from a reading point of view. I present first a brief outline of some basic features of the Southern Bantu languages. Agglutinating Languages: The Southern African Bantu Languages In the Southern African Bantu languages, all nouns belong to a noun class, which range in number from 14 to more than 23 (Schroeder, 2013). A noun prefix marks the class to which a noun belongs as well as marking number (singular/plural). Suffixes on nouns signal locative or diminutive functions. A concordial prefix (that is, a prefix that conveys agreement) ‘copies’ the class to which a noun belongs onto the verbal element of a sentence. The verbal element tends to be particularly packed with morphemes, with each morpheme carrying morphosyntactic information such as concordial agreement with the noun, and tense, aspect, and mood. The following sentence in Xhosa is an example: 1. Xhosa: Abafundi abayifundi – ‘The students (they) aren’t learning it’ Aba + fundi + ba + yi 66

noun class 2 (plural) prefix learnera negative marker noun class 2 concord 3rd person object concord

EARLY READING INSTRUCTION IN AFRICAN LANGUAGES

+ fund +i

verb stem negative marker

African agglutinating languages are syllable-timed languages (that is, languages in which pronouncing each syllable takes roughly the same amount of time) with a typical CV (consonant-vowel) or V syllable structure. Although they have a simple vowel system (between 5 and 7 vowels), there is a large and complex system of consonants represented by over 40 digraphs (e.g., hl, gx) and trigraphs (e.g., dlw, ndy), and several quadgraphs (e.g., ngqw, nkcw). Although there are prenasalized consonant structures such as -nq- or mf- that appear to be consonant clusters, these sounds are considered to be complex consonants representing a single phoneme and are therefore not proper consonant clusters. The double or triple consonant clusters of English CCV (brick; fleet) or CCCV/VCCC (street; thirst) do not occur in African languages. Instead, a word such as kwangentlazane ‘nearby’, has a CV+CV+CV+CV+CV structure (kwa-nge-ntla-za-ne). The orthographies of the Southern Bantu languages can be divided into two basic types: conjunctive and disjunctive. In conjunctive orthographies, the morphemes are written together so that they form a single written word unit; in a disjunctive orthography, some of the morphemes are written as separate word units. The languages belonging to the Sotho language group in South Africa and Botswana (Northern Sotho, Southern Sotho, Tswana) have a disjunctive orthography, while the Nguni languages (Xhosa, Zulu, Swati, and Ndebele) have a conjunctive orthography. Here is a sentence from each of these two groups along with the same sentence in English: 2. English: ‘The teachers aren’t reading books to the learners’ Northern Sotho1: Barutiši ga ba balele bana dipuku tša dikanegelo. Xhosa: Ootitshala ababafundeli iincwadi abantwana As the example shows, the conjunctive orthography of Xhosa produces dense text containing 11 morphemes written ‘together’ in four long words, while most of the morphemes in Northern Sotho are written separately as eight distinct words. To date, little research has been done on how best to teach reading literacy in African languages with a conjunctive orthography. Why is it important to be aware of these linguistic and orthographic differences? First, eye tracking and traditional reading research suggest that the profiles of readers in agglutinating languages may be different from those of readers in a language such as English. Despite sharing an alphabetic writing system, differences in the typology and orthography of a language may affect eye movements, the way in which words are recognized, and the role of phonological and morphological processes during reading. Eye movements that rest on words are termed fixations and last for about 200–300 milliseconds (ms) in fluent reading; the jumps between fixations are called saccades, lasting for about 30 ms (Rayner, 1998). Refixations are additional fixations made on a word before a reader leaves the word and moves to the next one. They may 67

PRETORIUS

reflect incomplete lexical processing of a word with the first fixation. In English, only about 15% of words in a text are refixated (Rayner, 1998, p. 387). Second, the developmental trajectories of readers of agglutinating African languages – especially those languages with a conjunctive orthography – may be different from those of English (e.g., Land, 2015; Probert, 2016; Van Rooy and Pretorius, 2013). Preliminary research, however, is only starting to emerge in this field. In the following section, I look at four orthographic features and consider their implications for early reading instruction in agglutinating African languages. These features are alphabetic transparency, the notion of a ‘word’, word length, and visual discrimination. Alphabetic Transparency Most of the world’s languages have alphabetic orthographies, where spoken language is represented at a sublexical level by letters of the alphabet (graphemes) that stand for, or represent, sounds of the language (phonemes). Sequences of letters then form words, which in turn form phrases and sentences. The colonial languages of Africa (English, French, and Portuguese) as well as all the Southern Bantu languages are all alphabetic languages. Alphabetic languages can be categorized on a continuum of greater or lesser transparency, depending on the consistency with which sounds map onto letters in written orthographies (Cook and Bassetti, 2005). Languages such as Spanish, German, Greek, Italian, and Welsh, which have a fairly clear one-to-one mapping between letter symbols and sounds, are called transparent orthographies. Languages with a more complex letter-sound orthography are called opaque orthographies. English is regarded as having an opaque orthography, with 26 letters representing about 42 to 44 phonemes. Much of the opaqueness in English orthography derives from its history of changing from an inflectional to a more analytic language, and from its complex vowel system. For example, English has about 20 vowel phonemes (depending on the variety or dialect of English) but only six letters to map them (a, e, i, o, u, and y). Thus, the letter ‘o’ is pronounced in eight different ways in words such as hot, love, so, do, or, bought, toy, how. French and Portuguese also have opaque orthographies. Agglutinating African languages have fairly transparent orthographies (Trudell and Schroeder, 2007). In most of the southern Bantu languages, for example, there is a simple vowel system of five to seven vowels represented by the five vowel letters of the alphabet. Many African languages are also, however, tonal languages, and tonal distinctions are not neatly captured in their alphabetic orthographies. This situation can create semantic ambiguity for a reader (Schroeder, 2013). Katz and Frost’s (1992) orthographic depth hypothesis proposed that the transparency of an orthography would determine the extent to which readers relied on lexical or phonological routes to decode text. Subsequent research with European languages seemed to bear this out. Readers of English, with its opaque orthography, seem to rely more on lexical routes when reading; that is, they recognize whole units 68

EARLY READING INSTRUCTION IN AFRICAN LANGUAGES

such as words and access them directly from the lexicon. By comparison, readers of transparent orthographies, such as Spanish, Italian, and German, access words in the lexicon via a phonological route, relying more on phonological decoding when they read than do readers of opaque orthographies (Cook and Bassetti, 2005). Ziegler and Goswami’s (2005) grain size theory posits that, to a large extent, the orthographic conventions of a language determine the size of the unit used while processing written language. ‘Grain size’ varies with orthographic mappings, ranging from phonemes, syllables, and morphemic units through to whole words. According to this theory, readers of opaque orthographies rely more on larger chunks (‘grains’) of letters such as rimes and whole words because letter-sound relations are not consistent (e.g., on English, Ziegler, Perry, Jacobs, and Braun, 2001; on Portuguese, Lima and Castro, 2009). By contrast, in transparent orthographies, where letters and sounds consistently correspond, a small grain, such as the phoneme, is used for decoding. Readers of these languages ‘assemble’ words, letter-by-letter, rather than chunking letters into larger units. Research into languages with transparent orthographies seems to bear this theory out. Researchers have looked, for example, DW*HUPDQ=LHJOHUHWDO :HOVK(OOLVDQG+RRSHU 7XUNLVK%DED\D÷LW and Stainthorp, 2007), and Basque (Acha, Laka, and Perea, 2010). There is, thus, converging evidence for the primacy of phonological processing in transparent orthographies for both inflectional and agglutinating languages. According to the transparency hypothesis (e.g., Wydell and Butterworth, 1999), the extent to which an orthography is transparent can affect the ease with which children learn to read in the early years. Based on cross-language comparisons, Seymour, Aro, and Erskine (2003) found that children learning to read in a transparent orthography can develop decoding skills within a year, while children learning to read in English need at least two and a half years. Ironically, the three dominant colonial languages in Africa – English, French, and Portuguese – all have opaque orthographies. Since African agglutinating languages have transparent orthographies, it follows that reading research relating to transparent languages would be more pertinent than research in opaque languages. An awareness that small grain sizes are important in transparent orthographies should alert us to the importance of phonics programmes that explicitly teach letter-sound correspondences for African children learning to read in their home languages. The Notion of a Word The notion of what constitutes a ‘word’ depends on the morphological typology of a language. What counts as a word can, in turn, have implications for early reading development and for what the eye focuses on during the reading process. Morphemes can be classified as free and bound. Free morphemes function like words and can stand alone, while bound morphemes are ‘bits’ that are attached to a larger unit. Thus, the verb love in English is a free morpheme, while the third-person 69

PRETORIUS

singular -s, past tense marker –ed, and the present participle marker -ing are suffixes and, hence, bound morphemes. They cannot stand on their own, but always come attached to a verb (loves, loved, loving). Given the rich and complex morphological structure of agglutinating languages that rely on affixation, words are multimorphemic. In such languages, the notion of a word is problematic since a whole sentence may be a single ‘word’ consisting of a combination of bound morphemes. In such languages, reading does not consist of reading a sequence of single, isolated words, but rather a sequence of bound morphemes and stems. For example, a simple sentence in English such as I love you is comprised of three separate short words (three free morphemes). Its equivalent in Xhosa is one orthographic unit comprising 13 letters: ndiyakuthanda. (i + present formative + object you + love). Here, the orthographic word encapsulates an entire sentence. In English, the letter sequences that make up the above sentence can be learned and recognized quickly as separate units; they are short, bounded by space on either side, and visually distinct. The rapid and automatic recognition of words in languages such as English, French, and Portuguese (that is, languages that are partially isolating and inflectional) is possible because their orthographies have a higher incidence of free morphemes. In contrast, such boundaries become visually blurred in an agglutinating language; it is not productive to learn these letter sequences as a chunk, since a small change somewhere in the linguistic unit (the beginning, the middle or the end) can change the meaning, as the following examples show: 3. 4. 5. 6. 7. 8.

Ndiyanithanda Ndiyayithanda Andiyazithandi Ndiyawathanda Ndakuthanda Bendikuthanda

I love you all I love it (e.g., a dog) I don’t like them (e.g., cakes) I love them (e.g., horses) I have loved you (e.g., for a long time) I loved you (e.g., it’s over now)

Although these six examples constitute six different orthographic ‘words’, because these words are morphologically and semantically dense, readers need to work their way through them. Although at a quick glance the Xhosa words above look fairly similar because they contain a high recurrence of similar strings of letters, they cannot, in fact, be recognized easily or quickly at a glance. A small change in one or two letters has semantic consequences, and paying attention to detail is important. If a reader has difficulty reading the letters from left to right quickly and accurately, then the subtle meaning differences are likely to be lost, resulting in slower reading, loss of accuracy, and loss of comprehension. Reading in such languages requires reading many unspaced, bound morphemes in sequence, with the morphemes occurring in longish phrases or as single sentences. Research on eye tracking has shown that unspaced text is difficult to read. Spaces between words are important in controlling eye movement and helping readers to decode texts (Rayner, 2009; Rayner, Fischer, and Pollatsek, 1998). If spaces between words are removed in 70

EARLY READING INSTRUCTION IN AFRICAN LANGUAGES

English, the reading rate slows down by 30–50%. Conversely, splitting apart multiunit German compounds was found to speed up reading (Rayner, 2009). Van Rooy and Pretorius (2013) used eye tracking data in their research on the differences and similarities that grade 4 learners showed in reading English and Zulu texts. Even though the grade 4 learners’ home language was Zulu, they read the English texts more than 100 letters per minute faster, on average, than they read the Zulu texts. Results also showed that fixations were much longer when reading Zulu text than when reading English text (554.5 ms in Zulu versus 468.8 ms in English), with a higher number of refixations occurring in Zulu. These differences suggested that the morphologically denser written words of Zulu required more cognitive effort and processing time. Land (2015) found that competent adult Zulu readers read more slowly in Zulu than in English, with average speeds of 1021 letters per minute in Zulu versus 1380 letters per minute in English. Land (2015) also found more fixations, which also lasted longer, among competent adult readers reading in Zulu compared to English. This is indicative of a slower, letter-for-letter processing employed when reading texts in agglutinating languages. Learners of English devote a lot of time to developing automatic word recognition skills. Such an approach may not, however, be productive for African languages, where the early development of accurate and efficient ‘small grain’ phonological processing at the phonemic and syllabic level may prove more beneficial for young readers. Length of Words The length of orthographic words is closely related to the notion of what constitutes a word in reading. English has hundreds of short, single-syllable words. Many of them are also high-frequency words that occur in everyday discourse, both function words (well, the, in, on, for, not, lot, that) and content words (dog, cat, girl, boy, ball, shoe, red, blue). In contrast, multisyllabic words are the norm in African languages; single syllable words, especially in languages with a conjunctive orthography, are rare. Although there are several two-syllable words in the conjunctive orthographies, they tend to be mainly verb stems, such as vala, vula, and jonga. It is difficult to avoid three- and four-syllable, and longer multisyllabic, words in agglutinating African languages, even in early reading material for young children. Thus, an early reader2 typically contains sentences such as those below (the number of letters per word unit is shown in parentheses): 9. Ndiyahamba ngomso. (10 + 6) I’m going tomorrow. Ndilungiselela uhambo. (14 + 6) I’m preparing for the journey. Kufuneka ndilungiselele nosana. (8 + 14 + 6) I must also prepare the baby. Nalo luyahamba. Ndinetikiti lohambo. (4 + 9 + 11 + 7) She, too, is going. I have a ticket for the journey. According to Björnsson (1983, in Land, 2015, p. 83), while the average length of words in newspaper text in English is 4.6 letters, and in Finnish, 7 letters, Land 71

PRETORIUS

(2015) found that the average length of words in Zulu newspaper texts ranged from 7.4 to 8.3 letters. What are the implications of longer orthographic words for reading? Shorter words in English can be read in a single fixation, but this is not always possible in agglutinating languages. Land found that skilled adult Zulu readers recognized only 25% of words instantly, in a single fixation; the rest required ‘cognitive work’ (Land, 2015). Furthermore, eye-tracking research has found that up to 25% of the short function words in English can be skipped, since they can be inferred from the preceding words (Raynor, 2009; White, 2008). This is not possible in agglutinating languages. Land found that competent readers reading in Zulu skip fewer than 1% of words. Given the longer word units in agglutinating languages, too much information is lost if small grain sizes are not processed. In fact, half a sentence can ‘disappear’ if a word is skipped. Word length also affects reading times slightly differently in transparent and opaque orthographies. For example, Ellis and Hooper (2001) found that word length affected reading times by 70% in Welsh, whereas this effect happened only 22% of the time in English. Norvig (2009, in Land, 2015, p. 161) found that words of up to nine letters (for example, colonel and although) are read in a single fixation in English, suggesting the application of lexically driven strategies. In contrast, Land found that fewer than 25% of words in Zulu were read in a single fixation, and these were mainly short words (kodwa, ‘but’; yonke, ‘all’). Longer words in transparent orthographies seem to involve longer reading times and refixations, suggesting letterto-sound decoding strategies. Thus, when reading in African languages, attention to orthographic details through phonemic processing is vitally important, since the phonemic processing also carries vital morphosyntactic information. Visual Patterns Abadzi (2006) argues that it is easier to develop automaticity in reading when orthographies have words that are visually dissimilar. Visual dissimilarity is common among English words. In contrast, the syllable-timed nature and complex allomorphy (that is, variation in a morpheme’s form) of agglutinating languages result in many letter sequences with visually similar CV patterns, as the examples below demonstrate. 10. 11. 12. 13. 14.

Andiyayithandi Andizokuthanda Ndizakuthanda Ndisezakuthanda Ndisezanithanda

I don’t love it I will not love you anymore I will love you I will still love you I will still love you all

In English, the words are separated by a space, and all the words are visually distinct. In contrast, the sentences in the Xhosa versions are comprised of a single complex word with no spacing between morphemes. At a glance, the words appear

72

EARLY READING INSTRUCTION IN AFRICAN LANGUAGES

visually quite similar – they share the similar syllable strings ndi, ni, se, ya, yi, zo, za, thand-. Land (2015, p. 131) reports that letter strings are repeated far more often in agglutinating African languages than in English. For example, there are four threeletter strings (the, and, -ing, -ent) that occur 10 times or more in English newspaper text; by contrast, in Zulu newspaper text, there are at least 30 three-letter strings that occur more than 10 times (Land, 2015, p. 132). This high rate of recurring syllables increases the visual similarity of morphemes within complex words; this in turn requires attention to detail and, hence, more cognitive work while reading. Meaning is lost if any of the bits within the linguistic unit are decoded inaccurately. Sometimes a reader may backtrack or regress from right to left to a previously read word or line and fixate on it anew. A regression can be triggered by encountering a new word or a low-frequency word, or if a reader does not understand a portion of text and skips back to check (Just and Carpenter, 1987). Land (2015) found that there were more regressions over spans of text in Zulu than in English. For example, competent English readers average one regression every ±46 characters, while competent Zulu readers average one regression every 24 characters, with a range of 29 to 18 characters (Land, 2015). This suggests that the visual similarity across longer word units requires more back-checking in agglutinating African orthographies. Overall, suggests Land (2015), the longer fixations in Zulu can be attributed to a combination of two factors: longer orthographic words, and the agglutinative language structure that results in fewer identical word forms that repeat and so become familiar. IMPLICATIONS FOR READING INSTRUCTION

The primacy of phonological processing in alphabetic orthographies, even for opaque languages such as English, has been well established over the decades (Adams, 1990; Chall, 1967; National Reading Panel, 2000; Stanovich, 2000). Even so, reading research across languages suggests that, in spite of many similarities in reading alphabetic orthographies, readers may also develop specific and nuanced reading processes unique to the orthographic conventions and linguistic structures of their languages. For example, there may be trade-offs in allocating cognitive resources and processing time during reading, depending on the transparency of a language’s orthography and its morphosyntactic complexity. Early reading instruction that is attuned to the linguistic and orthographic features of the languages in question may increase the chances of children’s getting reading right from the start of schooling. In this section, I identify some issues that have a bearing on early reading in the African languages, with regard to early instruction in general (particularly decoding), developing reading materials, establishing reading norms in the early grades, and teacher training.

73

PRETORIUS

Early Reading Instruction First, it might be more appropriate for reading practitioners in Africa to look to research relating to reading in transparent orthographies and agglutinating languages such as Finnish, Turkish, and Basque, rather than assuming that English instructional practices can readily transfer to African languages. The six points below identify aspects of best practice in transparent orthographies. They pinpoint some Englishderived approaches that are best avoided, some that should be approached with caution, and some existing practices in African reading classrooms that need to be reviewed and adapted in light of current reading research. Explicitly teaching the code. Research shows that a solid understanding of lettersound relationships is needed in agglutinating, transparent orthographies (Acha et DO  %DED\D÷LW DQG 6WDLQWKRUS  6LOYHQ 3RVNLSDUWD DQG 1LHPL   ,QUHVHDUFKRQHDUO\UHDGLQJLQ7XUNLVK%DED\D÷LWDQG6WDLQWKRUS UHDGLQJ accuracy was found to reach ceiling levels within a year of schooling if children knew their letter-sound relationships. This aspect of reading should be assessed regularly throughout grade 1 so that problems can be identified and rectified early, before children pass to the next grade without this foundational knowledge. A systematic and explicit phonics programme, unique to the particular vowel and consonant letter-sound relationships of African languages, needs to be developed. The sequencing of consonants in phonics instruction should be determined by the frequency of consonants unique to each particular African language, not by following the sequence in English phonics programmes. Whole language approaches to early reading instruction. Whole language approaches should be avoided in early African-language reading classrooms. The tendency in these approaches to underplay phonological processing, to deemphasize or embed phonics, to rely on the recognition of whole words, and to use contextual clues to decode words is simply inappropriate for African languages, especially those with a conjunctive orthography. The whole language approach to early reading instruction relies on the opaque orthography of English and flies in the face of converging reading evidence from both transparent and agglutinating languages. Because of the high incidence of short, high-frequency words in English, children can learn to recognize many words in English through whole language pedagogy. Even so, many thousands of words in English are decodable, and reading research indicates that children learning to read in English also benefit from phonemic awareness and phonics (Adams, 1990; Chapman, Tunmer, and Prochnow, 2001; Stanovich, 2000). To date, research indicates that developing fast and accurate phonological processing is vital when learning to read alphabetic texts, especially for transparent orthographies and agglutinating languages.

74

EARLY READING INSTRUCTION IN AFRICAN LANGUAGES

Morphological patterns. Even with well-designed phonics programmes in English that are effective in helping children crack the alphabetic code, caution is needed regarding their instructional generalizability. For example, onset-rime patterns (for example, -ake: bake, cake, fake, hake, lake, make, rake, sake, take, wake) abound in English, and form a rich source of rhyming words. Wylie and Durrell (1970) state that nearly 500 high-frequency words can be derived from 37 rime patterns in English. This makes for a highly productive reading method for English, but not for syllabic, agglutinating African languages with complex allomorphy. Speed and automaticity in decoding. ‘Small grain’ cognitive processing is required for reading in multisyllabic agglutinating languages. Given the higher visual similarity of syllables, longer words, and the morphological density of word units, developing speed and automaticity may be different when reading in these languages. This is an area that awaits further research. It is likely that daily reading practice and constant exposure to the reading of extended texts will be important. Grain size flexibility. Even though readers tend to rely on the grain sizes shaped by their orthography, this does not mean that other grain sizes are not available as processing units. Thus, English readers, especially those who receive a solid phonics grounding in the early years, will access both small and larger grain size units, depending on the decodable nature of the word, and they can use phonological processes to decode new words not encountered previously. Besides the primacy of phonological processing in transparent orthographies, it is possible that slightly larger grain sizes involving syllabic and morphological processing may also play a dominant role in agglutinating languages, as early reading in Finnish (Silven et al., 2004) and Xhosa (Probert, 2016) suggest. Probert (2016) found greater morphological awareness among young readers when reading conjunctive Xhosa texts, with longer word units, than among young readers with the disjunctive Tswana texts, with shorter word units, where phonological awareness seemed to suffice. Comparing beginner, intermediate, and advanced readers in Spanish and Basque, Acha et al. (2010) found initially slow decoding processing in children, while proficient adult readers could identify constituents within word units more rapidly. Syllabic teaching. A traditional way of teaching reading in African languages has been to teach syllable strings such as ba-, be-, bi-, bo-, bu; ma-, me-, mi-, mo-, mu. Because of the pervasive syllabic nature of African languages, this makes sense. However, given the importance of finer grain phonemic processing in transparent orthographies, it is also important for learners to be able to segment syllables into constituent phonemes and to blend sequences of syllables into words. Eye tracking research into grade 1 readers in Zulu and Northern Sotho found that most of the readers in each cohort had minimal decoding skills. Despite a year of instruction in chorusing strings of syllables in the ba-, be-, bi-, bo-, bu manner, the students had great difficulty decoding words in connected text and often could not get past the 75

PRETORIUS

first letter or syllable (Pretorius, 2015b). Wilsenach (2016) found that phonemic awareness rather than syllabic awareness was a stronger predictor of reading skill in Northern Sotho. Developing Reading Material The phonological and morphological nature of African languages has implications also for developing graded reading materials, and for designing assessment instruments. For example, when assessing phonemic awareness in English, it is common to include items that require children to delete a sound at the beginning, middle or end of a word (for example, book without /b/ is ook; book without /k/ is boo). In the Nguni languages, with their conjunctive orthography, all nouns are preceded by noun class prefixes comprised of a V or VCV, and almost all words end with vowels: ihagu: ubusi: unopopi: amadoda:

pig honey doll men

Including nouns in a phonemic awareness test requires thoughtful application, since deleting initial or final sounds will involve only vowels. Verb stems can be used instead, since they start with consonants – bona, jika, xola. Deleting consonants at the end of a word is tricky, since few end in consonants. Similarly, designing alphabetic friezes or flash cards in conjunctive orthographies requires a different approach. Because all nouns are preceded by noun class prefixes, the prefix can be printed in a different colour to distinguish it from the sound being illustrated, or the sound in question can be printed in bold: a b c d

iapile ubisi icici idada

apple milk earrings duck

Establishing Reading Norms Determining reading norms in African languages needs to derive from Africanlanguage reading data. Given the large differences in average word length between English and African languages with conjunctive orthographies, it is imperative that unique reading speed norms be determined for each language. For example, according to Hasbrouck and Tindal (2006), readers whose home language is English have average reading speeds of 123 words per minute by the end of grade 4. In Xhosa and Zulu, such reading speeds would be unrealistic, even for adult readers. Pretorius (2015a) found that good grade 4 Zulu readers who read accurately and with fluency were reading between 42 and 46 words per minute in Zulu. In contrast, reading at 76

EARLY READING INSTRUCTION IN AFRICAN LANGUAGES

this rate in English would characterize an at-risk reader. This clearly indicates that reading fluency norms in agglutinating African languages, especially in conjunctive orthographies, will be substantially different from those in English. Implications for Teacher Training and Development The colonial languages of English, French, and Portuguese are often taught in tandem with African languages in the early years and eventually become the language of learning and teaching in most African countries. In order to reduce misconceptions about reading that lead to ineffective teaching practices, teachers must understand basic differences between inflectional and agglutinating languages, between opaque and transparent orthographies, and between sounds that are similar and different in the relevant languages; they must also understand the reading methods most suited to developing skilled decoding in these languages. This kind of knowledge related to content of instruction should be included in pre-service and in-service teacher training programmes. Teachers can also be shown how to help learners develop a range of reading strategies that include both small and larger grain sizes for reading in both transparent and opaque orthographies. CONCLUDING COMMENTS

Teaching reading in the early grades in the 21st century is a complex juggling act for which teachers need to be well prepared. An early-grade class teacher needs to organize each day so as to cover numeracy and literacy, as well as other subjects required by the curriculum. In multilingual countries, this usually means that the teacher has to plan the teaching of reading in two different languages in the daily and weekly timetable. This is challenging at the best of times; it is especially challenging when teachers are not adequately trained for the task and when the broader socioeconomic and educational infrastructures do not properly support them. It is important for children from low-income schools in developing countries to be taught reading in ways that will effectively launch them on successful reading trajectories from the start of their schooling. In this chapter, I have argued that a foundational principle for early reading development is to take into account the ways in which African and Western colonial languages are linguistically and orthographically different, and to consider how these might affect best practice in early reading classrooms. The orthographic transparency and agglutinative nature of African languages should provide a principled framework for ‘reading’ research on the African continent and for guiding approaches to early reading instruction in these languages. Developing a deeper understanding of the linguistic and orthographic nature of African languages will enable education officers and teachers to better understand similarities and differences underlying reading development in agglutinating languages with transparent orthographies, on one hand, and in, the

77

PRETORIUS

colonial languages with opaque orthographies, on the other hand. We await further African reading research to provide more insights into teaching reading in ways that are linguistically and orthographically congruent with agglutinating languages. In the meantime, it is important to be aware that instructional practices that may work for inflectional languages with opaque orthographies may not be suited to syllabic, agglutinating African languages with transparent orthographies. To urge that attention be paid to the linguistic and orthographic nature of the languages in which reading is taught in no way minimizes the importance of all the other contextual factors that are critical in early reading development. Literacy develops within particular contexts, and some contexts seem to be far more enabling of literacy development than others. From converging evidence across research, we know that reading is a learned skill that must be explicitly taught and practiced (Wolf, 2008) and that developing phonological processing skills is important in the reading of alphabetic written languages. It takes several hundred hours of reading on a regular basis to become a fast, efficient, and skilled reader. The curriculum for early schooling across developing countries needs to prioritize reading per se and the time spent on reading. Daily exposure to written language and opportunities afforded children to practice and enjoy their newly acquired reading skills are critical factors in developing skilled readers – in any language (Abadzi, 2006). In sum, what an effective early reading programme for African languages might look like awaits further systematic investigation, informed by the nature of agglutinating languages and their transparent conjunctive/disjunctive orthographies (e.g., do syllable, phonemic, and morphological awareness play different roles in early reading in agglutinating languages [e.g., Xhosa or Northern Sotho] compared to inflectional languages [e.g., Spanish] or analytic languages [e.g., English]; how do reading speed and fluency relate to comprehension in conjunctive or disjunctive orthographies?), as well as by research into classroom practices in developing contexts that can be shown to effectively support literacy development (e.g., is the syllabic ba-, be-, bi-, bo-, bu approach effective, or does instruction need to include phonemic awareness within syllables?). NOTES 1

2

Xhosa and Northern Sotho are two of the nine African languages spoken in South Africa and used as languages of instruction in the early grades. The Molteno Institute for Language and Literacy developed the Vula Bula series of readers in all nine African languages in South Africa. The readers are graded according to level of difficulty, 1 to 6 stars, based on morphosyntactic difficulty and length of phrases/sentences. The example is taken from a 2-star reader (i.e., an easy reader).

REFERENCES Abadzi, H. (2006). Efficient learning for the poor. Washington, DC: World Bank. Acha, J., Laka, I., & Perea, M. (2010). Reading development in agglutinative languages: Evidence from beginning, intermediate and adult Basque readers. Journal of Experimental Child Psychology, 105, 359–375.

78

EARLY READING INSTRUCTION IN AFRICAN LANGUAGES Adams, M. (1990). Beginning to read: Thinking and learning about print. Cambridge, MA: MIT Press. %DED\D÷LW 6  6WDLQWKRUS 5   3UHOLWHUDWH SKRQRORJLFDO DZDUHQHVV DQG HDUO\ OLWHUDF\ VNLOOV LQ Turkish. Journal of Research in Reading, 30, 394–413. Chall, J. S. (1967). Learning to read: The great debate. New York, NY: McGraw-Hill. Chapman, J. W., Tunmer, W. E., & Prochnow, J. E. (2001). Does success in the Reading Recovery program depend on developing proficiency in phonological processing skills? A longitudinal study in a whole language instructional context. Scientific Studies of Reading, 5, 141–176. Cook, V., & Bassetti, B. (2005). An introduction to researching second language writing systems. In V. Cook & B. Bassetti (Eds.), Second language writing systems (pp. 1–67). Clevedon, UK: Multilingual Matters. Ellis, N. C., & Hooper, M. (2001). Why learning to read is easier in Welsh than in English: Orthographic transparency effects evinced with frequency matched tests. Applied Psycholinguistics, 22, 571–599. Hasbrouck, J., & Tindal, G. A. (2006). Oral reading fluency norms: A valuable assessment tool for reading teachers. The Reading Teacher, 5, 636–644. Johnson, E. R. (2012). Academic language and academic vocabulary. Sacramento, CA: Achievement for All. Just, M. A., & Carpenter, P. A. (1987). The psychology of reading and language comprehension. Boston, MA: Allyn and Bacon, Inc. Katz, L., & Frost, R. (1992). The reading process is different for different orthographies: The orthographic depth hypothesis. Advances in Psychology, 94, 67–84. Land, S. (2015). Reading isiZulu: Reading processes in an agglutinative language with a transparent orthography (Unpublished doctoral dissertation). University of KwaZulu-Natal, Pietermaritzburg, South Africa. Lima, C. F., & Castro, S. L. (2010). Reading strategies in orthographies of intermediate depth are flexible: Modulation of length effects in Portuguese. The European Journal of Cognitive Psychology, 22, 190– 215. National Reading Panel (2000). Teaching children to read: An evidence-based assessment of the scientific literature on reading and its implications for reading instruction. Washington, DC: US Government Printing Office. O’Sullivan, M. C. (2002). Reform implementation and the realities within which teachers work: A Namibian case study. Compare: A Journal of Comparative Education, 32, 219–237. Pretorius, E. J. (2015a). Failure to launch: Matching language policy with literacy accomplishment in South African schools. International Journal for the Sociology of Language, 234, 47–76. Pretorius, E. J. (2015b). Where are their eyes? What can we learn about early reading in African languages from eye tracking data? Paper presented at the Pan-African Reading for All Conference, University of Cape Town, September 2015. Pretorius, E. J., & Knoetze, H. (2012). The teachers’ book club: Broadening teachers’ knowledge and building self-confidence. Musaion, 31, 27–46. Probert, T. N. (2016). A comparative study of syllables and morphemes as literacy processing units in word recognition: IsiXhosa and Setswana (Unpublished master’s thesis). Rhodes University, Grahamstown, South Africa. Rayner, K. (1998). Eye movements in reading and information processing: 20 years of research. Psychological Bulletin, 124, 372–422. Rayner, K. (2009). Eye movements and attention in reading, scene perception and visual search. The Quarterly Journal of Experimental Psychology, 62, 1457–1506. Rayner, K., Fischer, M. H., & Pollatsek, A. (1998). Unspaced text interferes with both word identification and eye movement control. Vision Research, 38, 1129–1144. Schroeder, L. (2013). Teaching and assessing independent reading skills in multilingual African countries: Not as simple as ABC. In C. Benson & K. Kosonen (Eds.), Language issues in comparative education (pp. 245–264). Rotterdam: Sense. Seymour, P. H. K., Aro, M., & Erskine, J. M. (2003). Foundation literacy acquisition in European orthographies. British Journal of Psychology, 94, 143–174.

79

PRETORIUS Silven, M., Poskiparta, E., & Niemi, P. (2004). The odds of becoming a precocious reader of Finnish. Journal of Educational Psychology, 96, 152–164. Spaull, N. (2013). Poverty and privilege: Primary school inequality in South Africa. International Journal of Educational Development, 33, 436–1447. Stanovich, K. E. (2000). Progress in understanding reading: Scientific foundations and new frontiers. New York, NY: Guilford Press. Taylor, N., & Taylor, S. (2013). Teacher knowledge and professional habitus. In N. Taylor, S. van der Berg, & T. Mabogoane (Eds.), Creating effective schools (pp. 201–232). Cape Town: Pearson Education. Truddell, B., & Schroeder, L. (2007). Reading methodologies for African languages: Avoiding linguistic and pedagogic imperialism. Language, Culture and Curriculum, 20, 165–180. Van Rooy, B., & Pretorius, E. J. (2013). Is reading in an agglutinating language different from an analytic language? An analysis of Zulu and English reading based on eye movements. Southern African Linguistics and Language Studies, 31(3), 281–297. White, S. (2008). Eye movement control during reading: Effects of word frequency and orthographic familiarity. Journal of Experimental Psychology: Human Perception and Performance, 34, 205–223. Wilsenach, C. (2016). Identifying phonological processing deficits in Northern Sotho-speaking children: Use of non-word repetition as a language assessment tool in the South African context. South African Journal of Communication Disorders, 63(2), e1–e11. Wolf, M. (2008). Proust and the squid: The story and the science of the reading brain. Cambridge, UK: Icon. Wydell, T., & Butterworth, B. (1999). A case study of English-Japanese bilingual with monolingual dyslexia. Cognition, 70, 273–305. Wylie, R., & Durrell, D. (1970). Teaching vowels through phonograms. Elementary English, 47. Ziegler, J., & Goswami, U. (2005). Reading acquisition, developmental dyslexia and skilled reading across languages: A psycholinguistic grain size theory. Psychological Bulletin, 131, 3–29. Ziegler, J., Perry, C., Jacobs, A. M., & Braun, M. (2001). Identical words are read differently in different languages. Psychological Science, 12, 379–384.

80

ROBERT SAVAGE AND MARIE-FRANCE CÔTÉ

5. THE TEACHING OF READING AND WRITING IN SECOND- AND MULTI-LANGUAGE CONTEXTS Evidence from Canada

OVERVIEW

For some time, learning to speak and to read more than one language was a necessity only for some. Nowadays, however, bilingualism has become, if not a necessity, then at least a highly valued skill in many societies, for a variety of reasons (immigration, travel, international business, and so on) (Genesee, 2004; Kuo and Anderson, 2010). Consequently, bilingual education is becoming increasingly commonplace round the world. However, little is known still about biliteracy, and several authors concur that it has received insufficient attention (Snow, 2008; Verhoeven, 2011). In this chapter, we will review contemporary theory and research relating to dual language instruction in early literacy, with a focus on the Canadian context. The purpose of this review will be to identify key aspects of effective dual language instruction that could potentially be generalized to other educational contexts. INTRODUCTION

This chapter will be divided into four main sections. First, we conduct a brief overview of major theory, exploring influential conceptual models of common language processes and transfer of learning from one language to another, along with psychological evidence related to transfer. Second, we examine what is known about effective second-language instruction from the perspective of enacted teaching and how it relates to dual language policy. In the third section, we undertake a critical, evidence-based analysis of dual language teaching, drawing specifically from the historic and contemporary case of French Immersion in Canada. This programme has been widely viewed as a major success and, moreover, has been an influence on immersion intervention movements worldwide. The final section assesses the findings of this research, contextualizing it in cultural, linguistic, geographical, and empirical-theoretical terms. We explore the limitations but also the potential for the generalization of these findings. In our conclusion, we consider the factors that serve as the basis for policy and practice in the field of bilingual education, while identifying a number of questions that require

© JOINTLY BY IBE-UNESCO AND KONINKLIJKE BRILL NV, LEIDEN, 2019 DOI: 10.1163/9789004402379_006

SAVAGE & CÔTÉ

further research, in particular the possibility that effective teaching and learning practices may be common to L1 (that is, a first language) and L2 (a second language). THEORIES OF DUAL LANGUAGE INSTRUCTION

A significant proportion of children around the world now either learn to read a second language or learn to read in their second language (Bialystok, Luk, and Kwan, 2005; Cenoz, Genesee, and Gorter, 2013; Genesee, 2008; McBride-Chang, 2004; Shin and Kominski, 2010; UNESCO, 2003). Children enrolled in bilingual programmes have to learn to read and write two languages in the same time frame in which children in monolingual programmes learn only one language. It has also been argued that the quantity of knowledge that needs to be taught in both languages exceeds that which can actually be taught in school (Kohnen, Nickels, and Coltheart, 2010). This observation suggests that children transfer what they have learned in one language when learning a second one, at least to some extent. Two major hypotheses have been proposed to account for bilingual proficiency. According to the transfer hypothesis, students transfer reading-related skills in one language when learning to read in another language (Geva and Genesee, 2006). According to the common underlying proficiency hypothesis (CUP) (Cummins, 1981, 1998), some aspects of language and literacy development are based on underlying cognitive abilities that support development of certain reading-related skills in any language. The CUP hypothesis, like the transfer hypothesis, is supported by findings in the literature such as the close relationship between L1 and L2 abilities (Cummins, 2005). Several studies have demonstrated high degrees of correlation between L1 and L2 reading skills (Cummins, 1979, 2005; Verhoeven, 1994). Indeed, the crosslinguistic relationships between L1 and L2 literacy skills have been established in the literature for many skills, such as phonological awareness (e.g., Bialystok et al., 2005; Comeau, Cormier, Grandmaison, and Lacroix, 1999; Chiappe and Siegel, 1999; Kruk and Reynolds, 2012; Snow, 2008), decoding (e.g., Genesee, Geva, Dressler, and Kamil, 2008; Bialystok et al., 2005), and orthographic processing (Deacon, Chen, Luo, and Ramirez, 2013; Deacon, Wade-Woolley, and Kirby, 2009). Even though these correlations are usually stronger between more similar languages, correlations concerning specific linguistic skills (reading strategies) are still significant across languages that do not share many similarities, such as French and Chinese (Snow, 2008). Nonetheless, Cummins’ CUP hypothesis has been criticized on a number of grounds. Genesee (1984) points out that it neglects to also take social factors into account, instead of linguistic and cognitive factors alone. Verhoeven (1994) argues that the inclusivity and broad scope of the CUP hypothesis makes it hard to test empirically. Furthermore, because studies investigating cross-linguistic transfer are predominantly cross-sectional and correlational in design, causality cannot be demonstrated (Genesee et al., 2008; Koda, 2012; Snow, 2008). Additional, highquality randomized intervention studies are needed in order to provide causal proof 82

SECOND-LANGUAGE CONTEXT IN CANADA

supporting the existence of cross-linguistic transfer and thus of the CUP hypothesis. Further research is also required in order to clarify the directionality of connections between languages, as well as the concept of ‘common underlying proficiency’ inherent to Cummins’ model (Koda, 2012). Arguably, both greater conceptual clarity and empirical research are needed to elucidate the putative common processes (the non-language-specific competencies that support reading development across languages). In conceptual terms, this means more precisely defining the nature of cognitive processes, as distinct from language components or experience and knowledge. In empirical terms, it means establishing patterns of ‘transfer’ and ‘nontransfer’ across languages with studies designed to reveal underlying candidate processes and causal links. The aim will then be to determine whether these processes are indeed common or not. Empirical Evidence Related to Biliteracy A large body of empirical evidence links reading proficiency in a first language with reading proficiency in a second language. This finding is of particular relevance to the current emphasis on dual language educational systems and outcomes. Studies focusing on students who learn to read in L2 or in two languages have found significant and positive cross-linguistic correlations between overall reading ability and the component skills of reading, such as phonological awareness, knowledge of letter-sound relationships, and decoding (Geva and Genesee, 2006). The evidence clearly points to the fact that learning to read in one language facilitates reading acquisition in the other. Reading Development in Dual Language Contexts There is also evidence that children who are English language learners can learn such component skills as letter knowledge, phonological awareness and word decoding before they have mastered other aspects of L2 spoken language, which suggests a high specificity of transfer between languages (Gersten and Geva, 2003; Stuart, 2004). Other evidence for cross-linguistic facilitation comes from the now-welldocumented research on dynamic neurocognitive interactions in bilinguals during language acquisition and processing. This pattern is most strongly highlighted in experimental studies that have explored ‘lexical retrieval’ and ‘lexical processing’ (studies of the speed and resources involved in naming words with test groups of fluent adults, for example, by Dijkstra and vanHeuven, 2002), as well as in studies of ‘code switching’ in speech wherein the cognitive costs and advantages of switching between two familiar languages are explored (Paradis, Nicoladis, and Genesee, 2000), to name just two sets of researchers. Outside of research on immersion, there have been attempts to explore the impact of the structure of different languages on the skill sets involved in reading. The psycholinguistic grain size hypothesis (Ziegler and Goswami, 2005) predicts that 83

SAVAGE & CÔTÉ

in ‘deep’ orthographies such as English, characterized by imperfect grapheme-tophoneme correspondences, stronger onset-rime phonological awareness may be more important. (An example of onset-rime patterns in English: -ake: bake, cake, fake, lake, make, take, wake.) However, at least when it comes to the case of dual language instruction in French and English, it would appear that phonemic abilities are most important for the acquisition of both languages. Haigh, Savage, Erdos, and Genesee (2011) investigated the link between phoneme and onset-rime awareness and reading outcomes in children learning to read in a second language. Phoneme and onset-rime awareness tasks closely matched for extraneous task demands were administered in English and French in the spring of kindergarten to 98 Englishdominant children in French Immersion programmes. Regression analyses suggested that English phoneme manipulation was a significant predictor of both English and French reading outcomes after controlling for letter-name knowledge and word identification. French onset-rime knowledge measured in kindergarten accounted for significant variance in French reading outcome measures. These results support the hypothesis that a link exists between English phoneme manipulation in kindergarten and both English and French reading outcomes in Grade 2. As such, the results are inconsistent with grain size theory. Writing Development in Dual Language Contexts Writing in dual language contexts has been explored somewhat less often than reading but has been the focus of growing attention more recently. Studies on writing development in bilingual students have examined the metacognitive processes and knowledge involved in writing, such as planning, formation and revision, as well as cognitive-linguistic skills and knowledge that have been associated with second language acquisition. Evidence points to close interactions between process-based aspects such as sense of story (Francis, 2000), discourse structure (Davis, Carlisle, and Beeman, 1999), and other elements, such as sense of audience, and awareness of the recursive nature of writing (for a review, see Roca De Lorias, Murphy, and Marin, 2002). Research on the psycholinguistic aspects of writing suggests that cross-linguistic interactions exist with respect to orthography (Deacon et al., 2009), listening comprehension (Dufva and Voeten, 1999), the use of a varied vocabulary /DQDX]HDQG6QRZ DQGV\QWD['XUJXQR÷OX0LUDQG$ULxR0DUWL  There is also evidence that L2 students are more likely to draw upon L1 when writing in L2 if they are still in the early stages of L2 development, or when L2 skills are poorly developed, or when learners encounter unfamiliar words or constructions (Lanauze and Snow, 1989; Edelsky, 1982; Nathenson-Mejia, 1989). To extend this literature, Savage, Kozakewich, Genesee, Erdos, and Haigh (2016) examined cross-linguistic predictions of writing development across several grades in bilingually educated students using predictors based on the ‘simple view of reading’ (SVR) (Hoover and Gough, 1990). The SVR is an influential cognitive model that views reading comprehension as the product of proximal word ‘decoding’ skills and 84

SECOND-LANGUAGE CONTEXT IN CANADA

narrative-level linguistic comprehension skills. The study was an extension of two earlier research studies designed to examine predictive relationships between early oral language and pre-literacy skills in immersion students’ L1 (English) and later risk for language and reading difficulties in L2 (French) (Erdos, Genesee, Savage, and Haigh, 2010, 2014). The earlier studies had, respectively, revealed that early assessment of students’ L1 reading-related and oral language abilities were significant predictors of later individual differences in L2 abilities as well as risk for later L2 reading and language difficulties. Erdos et al. (2014) observe that letter knowledge and phonological abilities measured in kindergarten and Grade 1 predicted risk for word reading difficulties, and furthermore that language measures of sentence completion and morphological abilities predicted later risk for language difficulties. Similar findings concerning the importance of phonology and alphabetic knowledge in L2 word reading have been reported by Bourgoin (2014). Thus, empirically underpinned predictive measures related to the SVR were already available. In their study, Savage et al. (2016) examined whether decoding and linguistic comprehension abilities, such as defined by the SVR, accurately predicted from Grade 1 the Grade 6 writing performance of English-speaking children (n = 76) who were educated bilingually in both English, their first language, and French, a second language. The prediction was made from (1) English to English; (2) French to French; and (3) English to French. The results showed that both decoding and linguistic comprehension scores predicted writing accuracy but rarely predicted persuasive writing. Within the set of linguistic comprehension tests, one test, Formulating Sentences (the ability to form a grammatically acceptable sentence from a given word), was a strong consistent within- and between-language predictor of later writing accuracy. Practically speaking, these results suggest that early screening for later writing ability using measures of sentence formulation early in students’ schooling, in either their L1 or L2, allows teachers to identify needs and from there to differentiate instruction in the primary grades. Potentially, these practices may help prevent writing difficulties in either L1 or L2. Theoretically, the results suggest that there are quite robust correlations between reading-related abilities and writing abilities within the same language but also across languages. This finding adds to a growing body of research suggesting facilitative cross-linguistic relationships between bilinguals’ developing languages. THEORETICAL APPROACHES TO THE CONCEPT OF TRANSFER

For more than a century, research on the interactions between languages has drawn on the concept of ‘transfer’. In the field of education, effectively harnessing transfer represents one of the ultimate goals of teaching and learning and is aspired to by most educationalists (Day and Goldstone, 2012). However, transfer is viewed in many different and occasionally contradictory ways, both conceptually and empirically (Barnett and Ceci, 2002; Day and Goldstone, 2012; Schwartz, Bransford, and Sears, 2005). In fact, some authors argue that meaningful transfer is rare (Bransford, 85

SAVAGE & CÔTÉ

Brown, and Cocking, 2000; Haskell, 2001; Nokes, 2009; Perkins and Salomon, 2012), whereas others characterize it as common or ubiquitous (Chi and VanLehn, 2012; Dufresne, Mestre, Thaden-Koch, Gerace, and Leonard, 2005). Furthermore, definitions vary across authors and theories (Butler, 2010). Several taxonomies can also be found in the literature that distinguish between either the content that is being transferred or the context in which it is being transferred, or both (Barnett and Ceci, 2005; Haskell, 2001). While researchers generally agree that transfer is the use of previously acquired knowledge in new situations, there is little further consensus on the topic (Barnett and Ceci, 2002; Lobato, 2012). Theoretical Perspectives on Transfer Many different views of transfer are found in the literature. In the classical view, transfer is considered as the application of knowledge-elements learned in a given context or situation to a new context or situation (Bransford and Schwartz, 1999; Day and Goldstone, 2012). According to this view, the learning situation and the transfer situation must share common features in order for previously acquired knowledge to be used in a new situation (Thorndike and Woodworth, 1901). It follows that the more features the two situations have in common, the greater the probability of transfer. Transfer is thus often characterized in terms of ‘distance’, that is, the relative amount of similarities between the learning task and the transfer task. Transfer can thus be near or far, depending upon whether the tasks share many similarities or only a few (see Barnett and Ceci, 2002, for a classification of nine dimensions of far transfer). The classical view has been criticized for ignoring motivational and contextual aspects of transfer in both the definition and the assessment of transfer (for a review of the main criticisms, see Nokes-Malach and Mestre, 2013). More recent views of transfer such as the ‘preparation for future learning’ hypothesis (Bransford and Schwartz, 1999; Schwartz and Martin, 2004) and the ‘actor-oriented perspective’ (Lobato, 2003; 2012) have sought to broaden both the definition of transfer and the ways in which it is assessed. To go beyond the very narrow focus of traditional models, these analyses integrate motivational and environmental factors and promote more qualitative evaluation methods. For instance, the goal of the ‘actor-oriented’ view is not to assess if transfer occurs or not, but to investigate the ‘whats’, ‘hows’, and ‘whys’ of transfer. Nokes-Malach and Mestre’s (2013) ‘transfer as sense-making’ theory advocates an ecological view of transfer by integrating the motivational, social, and ecological aspects of a situation and relating them to the cognitive mechanisms at work. While their model is promising and incorporates aspects of different views of transfer, it has yet to be fully explored empirically (one approach is illustrated in Murray, 1999). Across theories, many factors have been considered to strongly impact transfer. Some of the most important of these are information retention (Butler, 2010), deep understanding and expertise in a domain (Nokes, 2009; Chi and VanLehn, 2012), cognitive load (van Merriënboer and Sweller, 2004), and metacognition 86

SECOND-LANGUAGE CONTEXT IN CANADA

(Bransford et al., 2000; De Corte, 2003). Students’ attitudes and motivations (Perkins and Salomon, 2012) as well as teaching methods (Schwartz, Chase, and Bransford, 2012) and wider cultural values concerning L1 and L2 acquisition also need to be considered as playing an important role in enabling transfer. Cross-Linguistic Transfer Despite the lack of theoretical and conceptual consensus surrounding the concept of transfer, the biliteracy literature seldom defines it in any detail, and researchers rarely explicitly acknowledge the controversial aspects related to its definition and measurement. In addition, theories and empirical results from the wider transfer literature are rarely used to interpret the results of studies of cross-linguistic transfer. The concept of cross-linguistic transfer itself has also been considered imprecisely defined and imprecisely assessed (Kuo and Anderson, 2010). In the biliteracy literature, transfer is generally viewed as the positive (or negative) influence of one language on another (Kuo and Anderson, 2010), and somewhat more precisely as the conscious or unconscious transfer of prior knowledge drawn from one language when using another. Koda (2009) recently summarized two points of view found in the literature regarding the definition of cross-linguistic transfer. First, transfer can be considered as the reliance on first-language knowledge when second-language knowledge is not sufficiently developed. According to Koda, this view implies that the object of transfer is linguistic knowledge (that is, a set of rules) and that transfer is expected to cease when second-language proficiency is attained. The second point of view Koda relates presumes that cross-linguistic transfer is the automatic activation of first-language knowledge, triggered by second-language input. Automaticity in L1 is thus a prerequisite for transfer in L2. From this perspective, it is assumed that formfunction relationships rather than a set of fixed rules are transferred and consequently, transfer is not expected to cease at any point in L2 development. These two views of transfer may not be irreconcilable. The cross-linguistic transfer of specific rules may occur conjointly with the transfer of form-function relationships. These may just be different forms of transfer, happening at different moments of L2 development. In short, the concept of transfer has been, and continues to be, the subject of many controversies in the scientific literature. Many definitions and views of the concept exist, and the cognitive processes and mechanisms underlying this phenomenon are still under investigation. Controversies around the concept of cross-linguistic transfer are somewhat less salient in the dual language and biliteracy literature. However, this does not mean that precise definitions are available. General descriptions of the concept are often used, but what cross-linguistic transfer is and how it occurs remain open questions (Koda, 2009). It should be added that definitions of this concept often ignore the motivational and ecological factors related to transfer. Future studies will thus need to address these questions in order to provide a stronger foundation to research in this field. 87

SAVAGE & CÔTÉ

EDUCATIONAL POLICY AND PRACTICE OF DUAL LANGUAGE INSTRUCTION

There has been much evidence to suggest that the dual language policy of French Immersion in Canada has been successful (Cummins, 2014a). Spanning several decades, many evaluations of the academic and language learning outcomes of alternative forms of bilingual education in Canada have been conducted, beginning with the widely publicized and analysed French-English immersion programmes in Montreal, Quebec, in 1965 (Lambert and Tucker, 1972). In the original Canadian early total immersion programmes, English-speaking students received all literacy and academic instruction in French from kindergarten to Grade 2 or Grade 3. In such programmes, English students’ L1 was only used as a medium of instruction starting in Grade 2 or 3, and then only for approximately one hour per day. The amount of English instruction was increased in successive grades so that by the end of elementary school approximately half of instruction was in French and half in English. Since the initial programmes, a number of variations have emerged. Some schools use a 50-50 model from the start, distributing teaching across English and French. Other schools alternate the language of instruction on a week-on-week basis. These different local models are often dictated by local demands rather than by theory or evidence. Johnson and Swain (1997) summarize eight core features shared by all immersion programmes: ‡ ‡ ‡ ‡ ‡ ‡ ‡ ‡

The L2 is a medium of instruction. The immersion curriculum parallels the local L1 curriculum. Pedagogical support exists for the L1. The programme aims for ‘additive bilingualism’ – the principle that bilingual expertise will enhance both L1 and L2 competence. Exposure to the L2 is largely confined to the classroom. Students enter with similar (and limited) levels of L2 proficiency. The teachers are bilingual. The classroom culture is that of the local L1 community.

In practice, not all of these features can be observed in schools with immersion programmes in the Canadian context. Among other variations, many teachers in such programmes are not bilingual, exposure to L2 exists in a variety of cultural forms, and L1 and L2 curricula differ quite significantly. However, the core feature is that the L2 is the medium of instruction. The Impact of Immersion Programmes How then have these programmes as a whole been evaluated? Unsurprisingly, much, although not all, research on these kinds of programmes has focused on learner outcomes in first language, second language, and academic achievement (for reviews, see Genesee, 1987, 2004; Genesee and Jared, 2008). These studies were 88

SECOND-LANGUAGE CONTEXT IN CANADA

motivated by concerns that teaching early literacy and academic content through a language that students have not mastered could result in diminished proficiency in L1 oral and literacy skills and lower academic achievement. These concerns were often predicated on an underlying assumption that bilinguals’ languages developed separately and that subject matter learned in one language was not accessible through the other language. Evaluations of immersion programmes undertaken both in Canada and, subsequently, around the world have not substantiated any of these assumptions or concerns. They consistently find that students in French Immersion attain levels of L1 proficiency and academic achievement the same as or often higher than comparable students in L1-only programmes. At the same time, these students acquire high degrees of functional proficiency in the L2, to a level often equivalent to native French speakers (Cummins, 1998; Genesee, 2004; Genesee and Jared, 2008). There have also been few reported differences in L1 English outcomes between programmes in which English instruction is delayed until Grade 3 (as in full immersion models) and programmes in which some English is introduced in the early grades (as in partial immersion models, Genesee and Jared, 2008). Careful analyses have also revealed few differences in outcomes between children of low socio-economic status, allophone children (those with a home language other than English or French), and children with low L1 language abilities (Genesee and Jared, 2008). On the other hand, studies consistently report that L2 French learners in immersion programmes consistently display relative weaknesses, compared to French L1 speakers, in expressive language skills, especially writing, vocabulary, and aspects of grammar (Cummins, 2014a). Several scholars have drawn attention to relative weaknesses in the practice of dual language programmes or the evidence base for them. For example, Cummins (2014a) launches a stinging critique of the misuse of the evidence base for dual language policy in Canada, arguing that the teaching of languages is separated into ‘two solitudes’ through the practice of using teachers who are proficient in only one language. In dual language programmes, for students to receive language instruction in one language, without reference to the other, flies in the face of evidence that L1 skills should be used to enhance L2 learning in an organic manner within and across classrooms. The Impact of Dual Language Instruction for At-Risk Students Research shows that immersion programmes have no negative effect on L1 literacy skills development in the long term (Slavin and Cheung, 2005). Thus, several studies have demonstrated that even though second-language learners in immersion programmes usually lag behind native-language learners in kindergarten in terms of their literacy skills, in most cases they catch up or, in some cases, even surpass the performances of native-language learners later in primary grades (Genesee, 2004; Lesaux, Koda, Siegle, and Shanahan, 2006; Lesaux and Siegel, 2003; Lesaux, Siegel, and Rupp, 2007; Slavin and Cheung, 2005; Verhoeven, 2000). However, several 89

SAVAGE & CÔTÉ

concerns have been expressed with regard to at-risk students (Genesee, 2007), those who have below-average academic performance, experience difficulties in their first language or have a learning disability (LD). Parents, educators, and researchers have questioned whether immersion programmes can be detrimental for at-risk students or if learning two languages can slow down development (Genesee, 2008; Genesee and Jared, 2008). There is also the question of whether differences emerge in comprehension or fluency on nuanced measures of these disabilities (Genesee and Jared, 2008; Erdos, Genesee, Savage, and Haigh, 2014). Unfortunately, to date, little evidence is available to answer these questions (Genesee, 2007; Kruk and Reynolds, 2012). One general finding that emerges from the research is that experiencing difficulties in L1 is a strong predictor of difficulties in L2. For example, Comeau, Cormier, Grandmaison, and Lacroix (1999) showed that students experiencing difficulties with phonological awareness in English experienced the same difficulties in French. Many other studies also demonstrated that early predictors of reading and language difficulty in L1, specifically phonological awareness and letter-sound knowledge (Erdos, Genesee, Savage, and Haigh, 2010; Genesee, Savage, Erdos, and Haigh, 2013; Jared, Cormier, Levy, and Wade-Woolley, 2010), can be used to predict reading difficulty in L2. In the same vein, Lesaux and Siegel (2003) demonstrated that the performance profile of immersion students experiencing difficulties learning to read as they learn in L2 (English) is approximately the same as that of nativelanguage learners. Both have marked difficulties in phonological awareness and working memory. Some empirical programme evaluation studies have suggested that at-risk students enrolled in immersion programmes perform better than at-risk students in full-time English programmes on comprehension measures (Kruk and Reynolds, 2012) and word reading measures (Da Fontoura and Siegel, 1995). Bruck (1985) compared the performances of students with academic difficulties in immersion programmes with those of students, also with academic difficulties, who had previously been enrolled in immersion programmes and who went back to full-time English programmes. Analyses revealed that students who switched programmes did not improve more than those who stayed in immersion. The research on below-average students suggests that being enrolled in an immersion or bilingual programme is not disadvantageous to English (L1) academic outcomes (Francis, Lesaux, and August, 2006; Snow, 2008). Genesee (1978) also found that below-average immersion students performed significantly better on French (L2) reading tests than belowaverage students in English-only programmes. Nonetheless, there have been few experimental intervention studies on French immersion students with learning difficulties in reading (Genesee, 2007; Genesee and Jared, 2008). Such studies are needed in order to learn more about the impact of immersion programmes and to guide the development of interventions for students enrolled in them. In their review, Genesee and Jared (2008) identified only three such research studies, and they note that only one had a control group and that none measured improvement in reading ability after the intervention. 90

SECOND-LANGUAGE CONTEXT IN CANADA

Future experimental intervention studies should include control conditions and direct measures of reading improvement to assess the effectiveness of language interventions. Empirical Research on French Immersion Pedagogy Cummins (1998) identifies two problem areas in the implementation of French Immersion programmes in Canada: the quality of French oral and written skills that students attain, and the relatively high dropout rate in some programmes across Canada. Cummins cites as part of the problem in both cases the methods and philosophy of teaching in some French immersion classrooms. Early observations of teacher-student interactions suggested an over-reliance on highly teachercentred or ‘transmission-oriented’ methods (Cummins, 1996; Cummins, 2014b). These approaches were seen as engendering minimal opportunities to use oral or written French for creative or problem-solving activities (Harley, 1991; Wilson and Connock, 1982). Compared to regular English language programmes, the immersion programs apparently used considerably less cooperative learning and project-based work, less creative writing (in French), and less reading of authentic children’s literature (Cummins, 1987, 1995). Quality of Classroom Teaching While pedagogical practices have substantially changed in many classrooms in the time since Cummins (1979) first drew attention to these issues, it is still possible to hear L2 teachers advocate for strong teacher-led instructional models, sometimes based on the fear that children will not otherwise pay attention in L2 classrooms. In addition, Cummins’ (2014b) speculation concerning the role of French Immersion pedagogy on learning outcomes, while important, has not been tested empirically. Thus, to evaluate the effects of instructional models and contexts on outcomes in French Immersion contexts, Savage and Pace (2019) investigated the relationships between the observed quality of classroom teaching and literacy outcomes, deliberately assessing relatively broad features of teaching that have been identified as important predictors of learning outcomes in monolingual contexts. These features were classroom Atmosphere (the consensual, democratic, emotionally focused nature of classrooms), Instruction (the balance between direct instructional strategy and student self-sufficiency), Management (the encouragement of student self-regulation of learning), and Student Engagement (students’ on-task behaviour). These points were assessed using a well-validated tool entitled AIMS (Roehrig and Christesen, 2010), and then related to growth in reading and listening comprehension between year beginning (t1) and year end (t2) in typical first language (L1) English and second language (L2) French learners exposed to simultaneous dual language instruction. A total of N = 332 children in 36 classrooms were followed through Grade 4, and were observed in the fall and winter semesters. 91

SAVAGE & CÔTÉ

The results of the hierarchical analyses showed that, with the exception of instruction in the fall semester (which tended to be more teacher-led in L2 classrooms), overall ratings of AIMS sub-scales across classrooms in L1 and L2 did not differ significantly across L1 and L2 contexts. This finding suggests that contemporary French Immersion teaching and learning is not generally characterized by systematically greater levels of teacher-centred ‘transmission’ modes of teaching in L2 classrooms compared to L1 classrooms. Further analyses revealed, however, that these overall classroom teaching quality ratings were not correlated across L1 and L2 classes within schools, showing that quite often children can be exposed to a teacher with a child-centred teaching philosophy in one language and a teacher with a teacher-centred philosophy in the other language. Such findings draw attention to the possible role of ‘systems’ or whole-school approaches for coherence in sharing teaching models. Finally, and importantly, the AIMS ratings over a one-year period predicted growth in classroom-level variance in reading and listening equally for L1 and L2. The latter results are important in that they suggest that broad and stable commonalities exist in effective L1 and L2 classrooms that can be readily observed in French Immersion classrooms. In other words, what we know broadly about good teaching in monolingual contexts (a focus on student-centred models of practice, balancing democratic and emotionally supportive learning, encouragement of student self-regulation alongside teacher modelling of strategies for learning) appears to apply equally in L2 contexts. Currently, considerable variation exists in these practices across L1 and L2 instructional contexts, even within the same schools, but not as Cummins (2014a) noted, systematically between L1 and L2 teachers, at least in the Quebecois context studied here. These results are thus amongst the first to look empirically at the bigger picture of teaching practices. It is clear, however, that there is a need for further research on optimal teaching practices in dual language contexts, and analysis of where they are similar to, and differ from, those in L1 or monolingual contexts. Best Practices in Dual Language Instruction Research on the specifics of dual language instruction has existed for some time and is diverse, multidisciplinary, and flourishing (August and Shanahan, 2006). For example, it has analysed the use of ‘cognates’ (words shared across languages), and the use of ‘bootstrapping’ skills in the application of L1 competencies in L2 comprehension and recommended both as important tools in language teaching (August and Shanahan, 2006; Cloud, Genesee, and Hamayan, 2000). Research exists on candidate micro-processes in literacy teaching practice. Lyster, Collins, and Ballinger (2009), for example, conducted a study based on a bilingual readaloud project in three classes of six- to eight-year-old children in which both English and French teachers read from a common storybook in each language. The study explored whether such practices raise teachers’ awareness of their students’ bilingual resources, and whether they engender both teacher and student cross-linguistic 92

SECOND-LANGUAGE CONTEXT IN CANADA

collaboration. The results of qualitative focus-group analyses pointed to enthusiasm for the project among both students and teachers. There was also enthusiasm for the collaborative approach used in the project, as well as the opportunities it created for both parties to focus on language and content. In a subsequent study, Lyster, Quiroga, and Ballinger (2013) investigated the effects of biliteracy instruction on morphological awareness in English and French through an intervention based on derivational morphology and illustrated books. The results suggested that measures of morphological ability improved in both languages following the intervention. Gersten and Geva (2003) document the results of classroom observations of young English language learners (ELLs), highlighting the following effective practices: (a) embedding phonological awareness learning within writing and decoding; (b) infusing lessons with systematic vocabulary learning; (c) defining an appropriate length to instructional activities while encouraging students’ responses and interactions so that even very young students express ideas in their own words; (d) using visual resources to convey meanings; (e) providing appropriate direct instruction of content and strategies for learning this content; (f) using explicit teaching of content at a level appropriate to ELLs in the sense that it allows high levels of participation, response accuracy, and opportunities for practice. Well-executed comparative studies of this kind provide useful recommendations for practices that can be adapted to French Immersion classrooms. However, further high-quality intervention research is needed, using optimal randomized control trial designs. Contrasts between intervention and control group outcomes in systematic pedagogical intervention studies in French Immersion classrooms would yield more conclusive findings concerning the causes of growth in learning and optimal teaching techniques in such L2 contexts (Genesee and Jared, 2008; Savage and Cloutier, 2017). A recent search for studies based on randomized control trial interventions in all available literature on dual language instruction (Côté, 2016) revealed only one such study, carried out in the United States. Goodrich, Lonigan, and Farver (2013) conducted a study with 94 preschool Spanish-speaking students learning English as an L2. Students were randomly assigned to either the control group or one of the two experimental conditions for a period of 21 weeks. The students in the control group followed the regular classroom curriculum. The students in the experimental condition received ‘pull-out’ instructional sessions, based on effective teaching practices related to phonological awareness and print knowledge. The researchers conducted four 20 minute sessions per week. In one of the experimental conditions, instruction was in English only (L2), while in the other the instruction was in Spanish (L1) for the first nine weeks and in English (L2) for the remaining weeks (transitional condition). Vocabulary, phonological awareness (phoneme blending and segmenting tasks), and print knowledge were measured before and after the interventions. At post-test, students in both experimental conditions scored higher than children in the control group on English (L2) measures. However, only children in the transitional condition scored higher than children in the control condition on Spanish measures (L1) at that point. 93

SAVAGE & CÔTÉ

This finding shows that the intervention in English alone was not sufficient to promote cross-linguistic transfer back to L1. The CUP hypothesis predicts that improvement in one language (either L1 or L2) should lead to improvement in the other language, but it was not the case here, at least at immediate post-test. The intervention in English led to an improvement in English (L2) but not in L1 (Spanish). It may be that L2 growth is somewhat less influential on L1 than the other way around (Cummins, 1998). The findings also highlight that children with greater ability in one language benefited more from instruction in the other language than children with lower skills, providing support for Cummins’ ‘threshold hypothesis’, which posits that transfer occurs only when a threshold level of ability is achieved in L1. Overall, this study suggests that the transfer of phonological awareness skills was caused by the intervention, but only for the most capable students. However, the fact that instruction was given in both languages in the transitional condition makes it very difficult to investigate cross-linguistic transfer per se. Clearly, with only one true experiment available in the literature, far more analytic studies and wellexecuted theory-driven intervention research remain to be undertaken on a range of language abilities to test questions relevant to best educational practices in dual language contexts. CONCLUSIONS: TO WHAT EXTENT CAN FINDINGS RELATING TO DUAL LANGUAGE INSTRUCTION IN CANADA BE GENERALIZED?

Beyond this chapter, the current volume seeks to explore language and literacy development in diverse contexts. To what extent, then, can the Canadian experience inform practice in other educational contexts, for instance in Asia or in East Africa? It is crucial to bear in mind that the current research was undertaken in a context that may differ greatly from other dual language instruction environments. For a host of historical and geopolitical reasons, it is important to explore this context and not simply export it uncritically to any other educational context that is likely to be differ culturally, economically and politically, among other respects. As we have shown, the research in the Canadian context highlights that immersion can be a successful enterprise. Indeed in many regards, it is often more successful than monolingual models of instruction. The impact of immersion appears particularly clearly for academic outcomes of receptive language such as language comprehension and reading. However, these findings do not guarantee that immersion will always be successful across all social contexts. That the official languages of English and French both have a European origin and are structurally similar in many important ways (semantically and phonologically) facilitates their cross-fertilization through instruction. The identification of ‘cognate’ words across the two languages, a useful pedagogical technique in Canada would not, for example, be applicable where one language is of European origin and the other of Asian origin. Both official languages in Canada 94

SECOND-LANGUAGE CONTEXT IN CANADA

have high status and are equally valued economically. Proficiency in either language is associated with economic success, and proficiency in both often engenders further economic benefits (see Heller, 1999, for further insightful discussion of this issue). Such conditions are not always present (the comparison with the status of Spanish in the United States context is instructive in this regard). While the structural similarity between languages can be overplayed as a predictor of immersion outcomes (Gersten and Geva, 2003), the cultural, economic, and motivational aspects of second language learning remain important factors in acquisition (Genesee and Jared, 2008). As to these: Canada does relatively well in international literacy comparisons such as PISA (Programme for International Student Assessment). PISA, a periodic formal assessment of literacy abilities in representative samples of adults, provides a picture of national literacy levels and has been used as the basis for international literacy comparisons. The relatively strong performance of Canada suggests generally strong L1 abilities in its citizens. Canada has the economic capacity to support learning with a well-developed universal educational system open to all citizens from ages six to 17 which is funded through taxation and without financial charge at the point of delivery. Finally, while we have shown that, in a general sense, immersion is successful in many respects, little evidence of a causal nature exists. Still lacking are wellexecuted and randomized intervention studies or evaluation studies to show whether particular pedagogical practices within French Immersion are differentially effective or whether particular styles of intervention teaching are efficient. It has also been argued that where relevant evidence does exist (on the highly connected nature of L1 and L2 learning), it has not always informed policy and practice (Cummins, 2014a). We have, however, drawn attention to recent work that gives at least some indication of the sorts of general classroom organizational features likely to be maximally effective. A dearth exists, also, of theoretically driven intervention research or work conceptualizing sufficiently well either transfer or its common underlying processes. While further research is undoubtedly necessary, one can reasonably assert on the basis of empirical evidence that French Immersion classrooms characterized by democracy, positive emotional tone, student responsibility, judicious balance between teacher instruction, modelling (particularly of strategies for learning), and student self-regulation (including small-group and peer-assisted work) encourage language use and thus facilitate L1 and L2 language and literacy acquisition. These broad features of dual language instruction, we argue, are shared with the best instructional practices in monolingual contexts. In this sense, they are perhaps ‘universal’. We would argue that the extent to which they can be seen as ‘universal’ across cultures in non-Western contexts, such as East Africa, is something that should be sensitively explored, while respecting local structures that may influence outcomes in ways unknown to us. Finally, in this vein, one key finding of 40 years of careful research on dual language English language learners and French Immersion instruction is the importance of harnessing skills in the first language (Cummins, 1998, 2014a). As August and 95

SAVAGE & CÔTÉ

Shanahan (2006) have argued, children are not ‘blank slates’ but rather bring key skills that can be used to develop their L2. L1 is thus a resource for learning, and, as we have pointed out (Genesee et al., 2013), for effective assessment, thereby constituting a key aspect of effective teaching. Such insights suggest that both L1 and L2 abilities can ultimately be enhanced in the best ‘immersion’ models. When effectively delivered, they serve to promote cultural and linguistic plurality. .

REFERENCES August, D., & Shanahan, T. (2006). Synthesis: Instruction and professional development. In D. August & T. Shanahan (Eds.), Report of the National Literacy Panel on language-minority children and youth (pp. 415–488). Mahwah, NJ: Lawrence Erlbaum. Barnett, S. M., & Ceci, S. J. (2002). When and where do we apply what we learn? A taxonomy for far transfer. Psychological Bulletin, 128(4), 612–637. doi:10.1037/0033-2909.128.4.612 Barnett, S. M., & Ceci, S. J. (2005). Reframing the evaluation of education: Assessing whether learning transfers beyond the classroom. In J. P. Mestre (Ed.), Transfer of learning: From a modern multidisciplinary perspective (pp. 295–312). Greenwich, CT: Information Age. Bialystok, E., Luk, G., & Kwan, E. (2005). Bilingualism, biliteracy, and learning to read: Interactions among languages and writing systems. Scientific Studies of Reading, 9(1), 43–61. doi:10.1207/ s1532799xssr09014 Bialystok, E., McBride-Chang, C., & Luk, G. (2005). Bilingualism, language proficiency, and learning to read in two writing systems. Journal of Educational Psychology, 97(4), 580–590. doi:10.1037/00220663.97.4.580 Bourgouin, R. (2014). The predictive effect of L1 and L2 early literacy indicators on reading in French immersion. The Canadian Modern Language Review, 70, 355–380. doi:10.3138/clmr.2346 Bransford, J. D., Brown, A. L., & Cocking, R. R. (2000). How people learn: Brain, mind, experience and school. Washington, DC: National Academy Press. Bransford, J. D., & Schwartz, D. L. (1999). Rethinking transfer: A simple proposal with multiple implications. Review of Research in Education, 24, 61–100. doi:10.3102/0091732X024001061 Bruck, M. (1985). Consequences of transfer out of early French immersion programs. Applied Psycholinguistics, 6(2), 101–120. doi:10.1017/S0142716400006068 Butler, A. C. (2010). Repeated testing produces superior transfer of learning relative to repeated studying. Journal of Experimental Psychology, 36(5), 1118–1133. doi:10.1037/a0019902 Cenoz, J., Genesee, F., & Gorter, D. (2013). Critical analysis of CLIL: Taking stock and looking forward. Applied Linguistics, 35(3), 243–262. doi:10.1093/applin/amt011 Chi, M. T., & VanLehn, K. A. (2012). Seeing deep structure from the interactions of surface features. Educational Psychologist, 47(3), 177–188. doi:10.1080/00461520.2012.695709 Chiappe, P., & Siegel, L. S. (1999). Phonological awareness and reading acquisition in Englishand Punjabi-speaking Canadian children. Journal of Educational Psychology, 91(1), 20–28. doi:10.1037/0022-0663.91.1.20 Cloud, N., Genesee, F., & Hamayan, E. (2000). Dual language instruction: A handbook for enriched education. Boston, MA: Heinle & Heinle. Comeau, L., Cormier, P., Grandmaison, É., & Lacroix, D. (1999). A longitudinal study of phonological processing skills in children learning to read in a second language. Journal of Educational Psychology, 91(1), 29–43. doi:10.1037/0022-0663.91.1.29 Côté, M.-F. (2016). Cross-linguistic transfer of literacy skills between English and French among grade 1 and 2 elementary school students (Unpublished doctoral dissertation). McGill University, Montreal. Cummins, J. (1979). Linguistic interdependence and the educational development of bilingual children. Review of Educational Research, 49(2), 222–251. doi:10.3102/00346543049002222 Cummins, J. (1981). The role of primary language development in promoting educational success for language minority students. In California Department of Education (Ed.), Schooling and language minority children: A theoretical framework (pp. 3–50). Los Angeles, CA: California State University.

96

SECOND-LANGUAGE CONTEXT IN CANADA Cummins, J. (1987). Bilingualism, language profiency, and metalinguistic development. In P. Homel & M. Palij (Eds.), Childhood bilingualism: Aspects of linguistic, cognitive, and social development. Hillsdale, NJ: Lawrence Erlbaum. Cummins, J. (1995). Empowering minority students: A framework for intervention. In O. Garcia & C. Baker (Eds.), Policy and practice in bilingual education: Extending the foundations (pp. 103–117). Philadelphia: Multilingual Matters. Cummins, J. (1996). Empowering minority students: A framework for intervention. In T. BeauboeufLafontant & D. S. Augustine (Eds.), Facing racism in education (2nd ed., pp. 349–368). Cambridge, MA: Harvard Educational Review. Cummins, J. (1998). Immersion education for the millennium: What have we learned from 30 years of research on second language immersion? Paper presented at the Second Katoh Gakuen International Symposium on Immersion and Bilingual Education, Katoh Gakuen, Japan. Cummins, J. (2005). Teaching for cross-language transfer in dual language education: Possibilities and pitfalls. Paper presented at the TESOL Symposium on Dual Language Education: Teaching and Learning in Two Languages in the EFL Setting, Bogazici University, Istanbul, Turkey. http://www.achievementseminars.com/seminar_series_2005_2006/readings/tesol.turkey.pdf Cummins, J. (2014a). To what extent are Canadian second language policies evidence-based? Reflections on the intersections of research and policy. Frontiers in Psychology, 5, 1–10. doi:10.3389/ fpsyg.2014.00358 Cummins, J. (2014b). Beyond language: Academic communication and student success. Linguistics and Education, 26, 145–154. http://dx.doi.org/10.1016/j.linged.2014.01.006 Da Fontoura, H. A., & Siegel, L. S. (1995). Reading, syntactic, and working memory skills of bilingual, Portuguese-English Canadian children. Reading and Writing: An Interdisciplinary Journal, 7(1), 139–153. doi:10.1007/BF01026951 Davis, L. H., Carlisle, J. F., & Beeman, M. (1999). Hispanic children’s writing in English and Spanish when English is the language of instruction. Yearbook of the National Reading Conference, 48, 238–248. Day, S. B., & Goldstone, R. L. (2012). The import of knowledge export: Connecting findings and theories of transfer of learning. Educational Psychologist, 47(3), 153–176. doi:10.1080/00461520.2012.696438 Deacon, H., Chen, X., Luo, Y., & Ramirez, G. (2013). Beyond language borders: Orthographic processing and word reading in Spanish-English bilinguals. Journal of Research in Reading, 36(1), 58–74. doi:10.1111/j.1467-9817.2011.01490.x Deacon, H., Wade-Woolley, L., & Kirby, J. R. (2009). Flexibility in young second-language learners: Examining the language specificity of orthographic processing. Journal of Research in Reading, 32(2), 215–229. doi:10.1111/j.1467-9817.2009.01392.x De Corte, E. (2003). Transfer as the productive use of acquired knowledge, skills, and motivations. Current Directions in Psychological Science, 12(4), 142–146. doi:10.1111/1467-8721.01250 Dijkstra, T., & vanHeuven, W. J. B. (2002). The architecture of the bilingual work recognition system: From identification to decision. Bilingualism: Language and Cognition, 5, 175–197. doi:10.1017/ S1366728902003012 Dufresne, R., Mestre, J. P., Thaden-Koch, T., Gerace, W., & Leonard, W. 2005. Knowledge representation and coordination in the transfer process. In J. P. Mestre (Ed.), Transfer of learning: From a modern multidisciplinary perspective (pp. 154–216). Greenwich, CT: Information Age. Dufva, M., & Voeten, M. J. (1999). Native language literacy and phonological memory as prerequisites for learning English as a foreign language. Applied psycholinguistics, 20(3), 329–348. 'XUJXQR÷OX$