ADDITIVE LANGUAGE LEARNING

FOR MULTILINGUAL SETTINGS

MAY 2014
This publication was produced for review by the United States Agency for International Development. It was
prepared by JBS International, Inc.
ADDITIVE LANGUAGE LEARNING
FOR MULTILINGUAL SETTINGS

DISCLAIMER
The author’s views expressed in this publication do not necessarily reflect the views of the United States
Agency for International Development or the United States Government.
 Additive Language Learning for Multilingual Settings

ACKNOWLEDGMENTS
This paper was written by Agatha J. van Ginkel, with support from JBS International, Inc.

i
 Additive Language Learning for Multilingual Settings

GLOSSARY
Affix is a morpheme (prefix, infix or suffix) that cannot stand on its own, but needs to be attached
to other morphemes/the root of a word.
Affixation is the process of using affixes to change words.
Agglutinative languages are languages in which there is a strong tendency for many affixes to be
attached to the root of words; because of that the languages have long words.
Bilingual education is education that uses two languages for instruction.
Derivational morphology is the process of creating a new word by changing an existing word.
When adding a morpheme to a word it does not only change meaning, but also word category. For
example in English ‘teach’ is a verb, but when adding –er ‘teacher’ it has become a noun.
Foreign language is a language that is not spoken in the immediate environment.
Mother Tongue is, for the purposes of this paper, the language spoken at the home of child and
the language a child knows best when he/she comes to school.
Multilingual education is education that uses more than two languages in instruction. In general, a
multilingual education program is a structured program that aims to develop cognitive, language, and
literacy skills in the first language and the additional language(s). It does so by making use of the
students’ first language as the medium of instruction for part of the curriculum.
Second language is a language that is spoken in the immediate environment and is also heard
outside school.
Script is the visual appearance of a writing system.
Writing system is a method of organizing the principles that guide how the symbols are mapped
onto the language unit.

ii
 Additive Language Learning for Multilingual Settings

CONTENTS

Acknowledgments ................................................................................................................................................... i
Glossary .................................................................................................................................................................... ii
Introduction............................................................................................................................................................. 1
Approaches to Combining Languages ............................................................................................................... 4
Bi/Multilingual Education and Transfer Issues.................................................................................................. 5
Thresholds and Competencies ........................................................................................................................... 6
Vocabulary Size as a Threshold....................................................................................................................... 6
Language Competencies ................................................................................................................................... 7
Case Studies from Different Countries ............................................................................................................ 9
Ethiopia ................................................................................................................................................................. 9
South Africa ....................................................................................................................................................... 10
Cameroon .......................................................................................................................................................... 11
Multilingual Classrooms ...................................................................................................................................... 13
Summary................................................................................................................................................................. 14
Recommendations ............................................................................................................................................... 15
Bibliography ........................................................................................................................................................... 18

iii
 Additive Language Learning for Multilingual Settings

INTRODUCTION
Most people around the world speak more than one language and many more would like to do so.
Africa, for example, is home to about one-third of the world’s languages, and most people there use
multiple languages to communicate in different domains and settings. Yet the rich linguistic
environment of most countries has been a challenge for policy makers and educators, who are faced
with difficult questions: what should a language policy for education look like? How do children learn
to read in different languages? What level of knowledge of a language is required before children
benefit from education using a new language as a medium of instruction? This paper aims to answer
this last question: when can the medium of instruction successfully include more than the mother
tongue? It also explores ways to determine this threshold level of knowledge and whether this
threshold level is different for different languages.
Countries have chosen different approaches to bilingualism and multilingualism. In general, however,
there are two main scenarios regarding language in education. The first is that a country has one
dominant language that almost every one speaks, but it wants to have multilingual citizens who can
interact freely in different languages and be global citizens. This is the situation in many countries in
Europe (Language Policy Unit, n.d.; Tragant, 2010). 1 The other is that a country has many languages
and a strong desire that every citizen learns one or two languages in addition to his or her mother
tongue so that the people can communicate with each other as a nation and be global citizens. This
exists in many countries in Africa, Asia, and Latin America. Moreover, in most of these countries the
language of the colonizer has left a strong footprint on people’s beliefs and perceptions about
language education.
Although both scenarios have the same goal of multilingual citizens, the process to achieve this goal
is different in each case. In the first scenario, there is often a strong tradition of teaching in a
language children understand—the mother tongue— throughout the school years. In Finland,
Denmark, and The Netherlands, for example, the mother tongue is the medium of instruction for
most students from primary school up to the university level. In fact, in several European countries
that have more than one native language, parents can choose which language they prefer for their
child as a medium of instruction. For instance, in Spain, primary education is provided in Spanish,
Basque and Catalan (Chenoz, 2008; Moreno, 2008). In the province of Friesland in The Netherlands,
parents can choose either the native language spoken there, Frisian, or Dutch as a medium of
instruction for primary and even secondary school (Gorter & van der Meer, 2008).In general using
the language students know as a medium of instruction has resulted in good education outcomes,
but, to a lesser extent, multilingual citizens (Fontecha, 2009). However, in the last few years, Europe
has made considerable progress in working toward multilingual—or what are sometime termed
"plurilingual"—citizens (Language Policy Unit, n.d.). This interest in plurilingualism has led to many
research projects aimed at gaining insight into how young students best learn other languages.
Bilingual schools are emerging, and additional languages are being piloted with young students (J.
Enever, Moon, & Raman, 2009). Also, in the last few years, university courses are increasingly
offered in English, rather than just the dominant native language. Interestingly, the countries that
produce the most proficient English language speakers are countries that do not use English as
medium of instruction in school (Education First, 2013).
In multilingual countries that follow the second scenario, children leave the mother tongue as soon
as possible and continue education in the official language—often the language inherited from the
colonizer—even though they do not speak this language when they enter school, nor is it spoken at
home. Consequently, most children in Africa find themselves having to try to learn through a
language they do not understand (Alidou, Boly, Brock-utne, & Satina, 2006; Djité, 2008; J. Enever et
al., 2008; Ouane & Glanz, 2006), and many of them fail in education.

1. This paper discusses several terms that relate to bilingual and multilingual education. The focus is, in particular, on learning to read and
write in bilingual and multilingual education (bi/multilingual education) settings.

1
 Additive Language Learning for Multilingual Settings

In several African countries, the mother tongue is recognized as having a place, particularly at the
policy level, but strong assumptions about the colonial languages—made by parents, teachers, and
education officials alike—hinder the implementation and use of the mother tongue in education. The
language of the colonizer is often the language used as a medium of instruction in education. In
Africa, this language is frequently seen as the language of success and socio-economic development
(Djite, 2008). “Many parents, especially in rural areas, express the view that they send their children
to school to learn the official language (English, French, or Portuguese), and that a policy forcing
them to learn an African language amounts to a dumbing down of their children and keep them at
the bottom of the socio-economic scale” (Djite, 2008, p. 21). One rather persistent assumption is
that the earlier children start to learn the colonial language, the better they will be at it, and that
longer learning of the language will give higher results (Alidou, Boly, Brock-utne, & Satina, 2006).
However, this assumption is not well supported by research conducted in Europe or Africa (Alidou
et al., 2006; J. Enever et al., 2009; Heugh, Benson, Berhanu, & Yohannes, Mekonnen, 2007). In some
specific contexts, such as in Canada, this submersion approach in which only the new language is
used has worked well. However, using English as a medium of instruction in settings where it is only
used in school does not result in good learning outcomes for the English language or other subjects
(Heugh et al., 2007).
Research from primary schools in Africa, where English is used as medium of instruction (MOI) in
the early grades but is barely spoken if at all outside the classroom, shows that very little learning is
taking place (Piper, 2010; Uwezo, 2010). The Early Grade Reading Assessment (EGRA) carried out in
Kenya shows this in greater detail. (See Figure 1.) Kenya has a policy stating that the mother tongue
should be used as the MOI in classes 1-3. The reality is that teachers use English most of the time,
53.9 percent, as compared to 27.7 percent Kiswahili, a national language, and 18.2 percent in the
mother tongue (Piper, 2010). Yet children scored only 7.8 percent on the reading comprehension
test in English; they scored more than twice as well—16.9 percent—in Kiswahili, and almost five
times better—38.2 percent—in the mother tongue. In most of the cases, children were not even
taught how to read in their mother tongue, yet they were able to transfer what they had learned in
Kiswahili and English to the language they knew best, the language they learned at home.
Figure 1. EGRA Results Rural Central Province, Kenya (Piper, 2010)
60
53.9

50

38.2
40

30 27.7

16.9 18.2
20

10 7.8

0
English Kiswahili Mother Tongue

MOI Reading comprehension

Additional research shows that starting education in a language children do not understand does not
yield high learning results, nor does it help children to learn the new language well (Alidou et al.,
2006; Mothibeli, 2005). Research from Europe shows that young students who had 3 to 4 years of
language learning did not reach a sufficient level of language skill for teachers to be able to use the
added language as an MOI in the classroom (J. Enever et al., 2009).

2
 Additive Language Learning for Multilingual Settings

However, reading comprehension data from Eritrea shows that children who learn to read in a
foreign language take 5 or 6 years to reach the same level of reading comprehension that children
who learn in the mother tongue reach in 3 years or less (S. L. Walter, n.d.). Research from a
bilingual education program in Cameroon shows that students who learn in their mother tongue and
learn English as a subject perform better on English language tests than children who learn in English
all the time (Walter & Trammell, 2010). Similar results are reported from South Africa (Taylor &
Coetzee, 2013). These data show that teaching in a language that students do not understand or
have not learned well yields poor cognitive results; the students lag behind their peers who receive
education through a MOI that they understand.
It is commonly believed that the earlier children start to learn a language, the better they learn it.
However, research from countries in Europe revealed that young students who started to learn
English at 10–11 years of age progressed more in two years’ time than very young 4-6 year old
students (J. Enever et al., 2009; Muñoz, 2008; Nikolov, 2009). Older children appear to perform
better in instructional settings because of their cognitive maturity. Also, instruction that is only a few
hours per week is more suitable for older learners. Teaching a foreign language to students who are
9 to 11 years old, therefore, is more productive. However, very young learners have a few
advantages: they acquire the sound system of the language more easily and they are less anxious
about learning a new language. Further, when children start at a very early age, language learning is
largely an intuitive process; as they grow older, processing becomes more analytical. The two
learning processes combine to help the new language become deeply embedded in their brains. Also,
learning an additional language has a positive influence on the general educational development of the
learners (Johnstone, 2002).
Certain key conditions should be in place for language learning to be effective with very young
learners. The conditions are contextually bound, but examples include (Djigunovich & Vilke, 2000):
• Intensive interaction in class;
• Instruction for 45 minutes per day, five days per week;
• Class size of 10–15 children;
• Teachers who have a good command of the language.

Identifying key context conditions is critical. “…if any of the requirements are missing, second
language instruction should not begin at an early age; a negative experience may harm children's
attitude to the target language and to language learning in general”(Nikolov, 2000, p. 43). Older
children, 10 years of age or more, have some learning advantages also. They can use concepts
learned in their first language (L1) to "plot" the new language; the very young learners still have to
learn those concepts. Moreover, older learners are more experienced in different language activities
and may gain more from feedback and negotiating meaning. Having acquired a wider range of
strategies, they are more efficient learners, have a better understanding of why they are learning an
additional language, and may be more purposeful in working towards their own objectives
(Johnstone, 2002).

The research suggests that the advantages different age groups have when learning new languages
should be considered and utilized. For example, if a language is to be taught in a classroom setting, it
may be better to start the added language after students have had 3 or 4 years of primary school.
These students will take less time to learn as much of the new language as learners who started in
their first year of primary school. Additionally, transitioning to using the new language as MOI is not
a must. In fact, there is sufficient evidence that using the added language as MOI too early may even
be detrimental to language learning and other learning.
The key question is how to best use language to optimize learning for children living in environments
where more than one language is spoken and where it is necessary to be bi/multilingual citizens.
Ignoring the languages that children speak at home and expecting them to come to school with
inherent knowledge of the language(s) of instruction is not an option.

3
 Additive Language Learning for Multilingual Settings

APPROACHES TO COMBINING LANGUAGES

There are three main approaches to combining languages in bi/multilingual education: the subtractive
or submersion approach, the transitional approach, and the additive approach (Alidou et al., 2006;
Bot & Herder, 2008; Unsworth, Bot, Persson, & Prins, 2012). The underlying philosophies of these
approaches express different messages about the value and importance of the languages spoken by
people in bi/multilingual settings.
The goal of the subtractive or submersion approach is to enable the learner to acquire the official
language as the medium of instruction as soon as possible. The first language is hardly present in
education and is not maintained. The child’s background, linguistic heritage, and culture are not
acknowledged. The submersion approach assumes "the earlier the better" and "the longer the
better."
The goal of the transitional approach is also to enable the learner to acquire the official language,
but the learner's first language is used for some years during primary education. There is a gradual
transition in which the child's mother tongue is left behind and a new language—often a foreign,
former colonial language—replaces it. This approach is also referred to as subtractive bilingualism
(Cummins 1994). The second language is added at the expense of the first language. This is the
approach in many countries in Africa, Asia, and Latin America.
Finally, the goal of the additive approach is that students have high proficiency levels in the first
language and the other language. Both languages are used in education, and the first language
continues to be developed. A second (and often a third) language is added as an enrichment of the
first one. This approach to multilingualism values the language of each individual. It respects the
individual’s background, linguistic heritage, and culture. It helps children to develop high proficiency
levels in their mother tongue and the language of wider communication.2 This is the approach in
most European countries.
A closer examination of these approaches provides important information about the contexts where
they have been used, when success was achieved, and what factors led to success in which contexts.
The choice of approach also communicates a message about the value of an individual’s linguistic and
cultural heritage. Such messages have long term consequences for people's self-esteem and the
cultural and linguistic diversity of nations.

2. For a more in-depth discussion, see Alidou et al., 2006.

4
 Additive Language Learning for Multilingual Settings

BI/MULTILINGUAL EDUCATION AND TRANSFER ISSUES

In bi/multilingual education, two or more languages interact in a person’s mind, influencing each
other in positive and negative ways. The languages are not separate units, but interlinked in the
brain, where transfer from one language to the other can take place (Cook, 2003). Cummins (1979)
postulated the linguistic interdependence hypothesis, which states that in bilingual development,
language and literacy skills can be transferred from one language to another.
Research with young bi/multilingual students has shown that reading acquisition in the new language
uses similar underlying skills and strategies as in the first language (Arab-Moghaddam & Sénéchal,
2001). Having developed phonological awareness in their first language helps students to transfer
that skill to the new language(s). However, this implies that young students need to have sufficient
knowledge of the new language to hear syllables and individual sounds. Clearly, the transfer of
reading skills and strategies from first language to the new languages depends on the student's
reading ability in the first language and language ability in the new language. Research with young
bi/multilingual students learning to read a new language also shows that oral language competencies
aid decoding and word recognition (Bossers, 1991; Carrell, Devine, & Eskey, 1998). The extent or
ease with which the transfer of reading skills and strategies takes place depends on the language and
writing system. For instance, for most students in Africa, the European languages used in education
are very different from their mother tongues in grammar, vocabulary (no cognates), concepts,
writing systems, and scripts. However, for most students in Europe, the linguistic distance between
their mother tongue and English is not that far. There are similarities in concepts, vocabulary, writing
system and script, and positive transfer can take place. Linguistic distance between the known
language(s) and the new language influences the speed with which new languages are acquired
(Chiswick & Miller, 2004).
The differences between writing systems and scripts also influence transfer. A writing system
consists of the principles that guide how symbols are mapped onto language units, and script refers
to the visual appearance of a writing system (Piper & Van Ginkel, 2014). Conceptually, writing
system and script are independent from each other, and they influence the reading process
differently (Almabruk, Paterson, McGowan, & Jordan, 2011; Coderre, Filippi, Newhouse, & Dumas,
2008; Coulmas, 1996; Nag, 2007). Learning to read in one language with a certain writing system and
script takes a different trajectory from learning to read in another (Asfaha, Kurvers, & Kroon, 2009;
Nag, 2007; Seymour, Aro, & Erskine, 2003). This means that for any additional language learned,
students need to acquire the writing system of that language and often also a new script. For
example, in Ethiopia, students who learn in alphabetic script languages, which are phoneme based,
need to learn Amharic, which has an alphasyllabic script based on syllables and more than 240
symbols. These students likewise need to learn the writing system of English, which is much less
predictable than their own mother tongues, which often have regular sound and symbol
relationships (Piper & Van Ginkel, 2014).
Research has also shown that transfer of reading skills and strategies between languages is influenced
by different linguistic variables including the transparency of the writing system, and its syllable
structure, word length, symbols, phonological unit, agglutination, and word frequency (Duibhir &
Cummins, 2012; Seymour et al., 2003; Van Ginkel, 2008; Ziegler et al., 2010). This knowledge about
learning to read in a bi/multilingual setting implies that there is not one method or approach that
could work for all contexts.
Besides the linguistic elements, other variables such as the language level of teachers, the motivation
of young students, the method used, the hours per week spent on language instruction and learning,
the amount of exposure to the new language outside the classroom, and the perceived value of the
additional language and the first language all play important roles (Bot & Herder, 2008; Enever et al.,
2009; Marianne Nikolov, 2006; Thijs, Trimbos, Tuin, Bodde, & de Graaff, 2011; Unsworth et al.,
2012). Again, this shows the complexity of bilingual/multilingual education and the importance of
understanding the context in which bi/multilingual education takes place.

5
 Additive Language Learning for Multilingual Settings

THRESHOLDS AND COMPETENCIES

One of the most urgent issues for policy makers and educators in multilingual settings is how to
know when students have enough knowledge of an added language to benefit from it as a medium of
instruction in education. What is the threshold level for a given language? While policy makers
would prefer to know how many years of education it would take to reach this level, to express this
level in years of education would be misleading. It would assume that the context for each child,
class, and education system was the same. But contexts differ in many ways: the use of the foreign
language outside of the school setting, the language level of the teachers, the materials available for
language learning, the number of students in the class, the hours of language learning, etc. (Enever et
al., 2009; Language Policy Unit, n.d.; Unsworth et al., 2012). A better way to express the threshold
level would be to know when language users have achieved sufficient skills so that education could
start to take place in the other language. It would also be useful to know what skills are contained in
that threshold level. In 1979, Cummins presented this language threshold as a hypothesis, which is
now known as the linguistic threshold hypothesis. The linguistic threshold hypothesis states that a
linguistic threshold is a necessary point that a bilingual student must reach in order to benefit from
bilingualism (Cummins, 1979). There are two ways of looking at the linguistic threshold level:
knowledge level or user competence.
VOCABULARY SIZE AS A THRESHOLD
To take knowledge level as a threshold means looking at what language knowledge (linguistic
knowledge) users need to know to be able to do well in the classroom. Research shows that limited
language proficiency hinders transfer of reading and other skills from the first language to the new
language (Bossers, 1991; Carrell et al., 1998). Of late, applied linguists have moved from looking at
grammar to looking at vocabulary because research has shown that the language threshold for
reading is largely based on the vocabulary size of the language user (Eyckmans, 2004; Hirsh & Nation,
1992; Nation, 2006). Before successful reading can take place in the new language, sufficient
knowledge of words in that language needs to be present. The question is how large a vocabulary is
necessary in order to begin to read and write in the other language?
Vocabulary size is often measured in word families in which the base word and all its inflected and
derived forms count as one (Nation, 2006). This assumes that, when reading and listening, a learner
who knows at least one member of a word family could understand other members by using
knowledge of the most common and regular English word-building devices (Nation, 2006). However,
other studies have shown that it is important for English language students to also learn derivational
morphology—the process of forming a new word on the basis of an existing word—in order to be
more successful (Kieffer & Lesaux, 2007).
Nation (2006) recommends that language students learn high-frequency and academic vocabulary. In
the early stages of language learning, this is particularly important because students' language use
depends heavily on the number of words they know. Nation states that a significant threshold in
English is about 2,000 word families. When students know fewer words, the materials they use need
to be simplified and graded at their level. Graded readers use limited vocabulary to support the early
language students; the students encounter the words often and learn the words better. Nation and
Wang (1999) suggest that about 10 repetitions of words are necessary for students to retain the
word and its meaning, but more repetitions are better. Teaching the most frequent words to young
students is also helpful for languages other than English. Languages have different vocabulary sizes for
threshold levels; they also have varied affixation and word length, all of which affects frequency
calculations. Agglutinative languages will need a different calculation, probably also including frequent
morphemes. Research is starting to confirm that languages indeed differ in how many words need to
be learned to reach a threshold. The threshold level for French as a foreign language seems to be
about 2,200 frequent words, while the threshold level for Greek as a foreign language is about 3,350
frequent words (Milton, 2001).

6
 Additive Language Learning for Multilingual Settings

There is a strong link between vocabulary knowledge and reading comprehension (Milton, 2001;
Nation, 2006).To read with reasonable ease means that at least 95 percent of the words in a text
are known. When 97 to 98 percent of the words are known, unassisted comprehension can take
place (Hirsh & Nation, 1992; Nation, 2006). The 2,000 most frequent words in English cover about
80 percent of all text, which means that unassisted comprehension is still quite difficult when only
knowing these 2,000 words. Researchers tend to agree that, in English, a vocabulary of about 3,000
to 5,000 words is necessary for students to read authentic text (Eyckmans, 2004). This suggests that
for an English language learner to benefit from reading non-graded materials he needs to learn at
least 3,000+ words, including the 2,000 most frequent ones. It seems that knowledge of the most
frequent words in a new language is crucial for comprehension to take place in English and other
languages (Milton, 2001).
A question of interest is how long does it take English language learners to learn 3,000 or more
words so that they can read non-graded materials? While students whose mother tongue is English
have acquired about 4,000 to 5,000 word families by the age of 5 and learn about 1,000 new word
families each year, a realistic target for children learning English as a foreign language might be
around 500 word families a year, given good learning conditions. Nation (1990) describes studies
showing that after 5 years of regular lessons, children in India and Indonesia had between 1,000 to
2,000 word families. In Hungary, young students learning English as a foreign language learn about
1,460 words after 4 years of education (about 4 words per contact hour), while in Greece young
students learned about 2,280 words in 4 years (about 5–6 words per contact hour) (Orosz, 2009).
This seems to indicate that it would take more than 5 years of education for children to reach a
level of vocabulary knowledge that would allow them to use the English language comfortably and to
understand the content of their school books and teachers’ instruction.
Nevertheless, vocabulary size as a threshold has its limitations. Although it might be effective to
teach the most frequently used words in the early stages of language learning, the actual benefit for
learners will depend on the task, domain, and context in which the words will be used. With young
students, it is important to teach the words that are needed for learning new things in class and for
following instructions, not just everyday words. Furthermore, just teaching vocabulary to reach the
critical threshold is not sufficient. Young language students need to learn strategies that enable them
to learn more words by themselves in order to be prepared for a change in the medium of
instruction. They need to be able to figure out the meaning of words they do not yet know.
Additionally, knowledge of the structure of the new language and how text is built in the language
are crucial in transferring reading comprehension skills (C. Walter, 2004).
Thus, to only use vocabulary size as a threshold measure would not be sufficient. More than
linguistic knowledge is necessary for learning to be successful.
LANGUAGE COMPETENCIES
Another way to look at language learning is to look at what students are able to do with the new
language. In different settings, assessment scales have been developed to measure language learning
based on competency. The United Kingdom uses the Common Scale for English as Additional
Language or EAL (Qualifications and Curriculum Authority, 2000), while the European Union has
developed a framework that can be applied to all languages in Europe called the Common European
Framework Reference for Languages (CEFR) (Language Policy Unit, n.d.; Little, Goullier, & Hughes,
2011). The CEFR has become an influential instrument in language teaching and learning in Europe
and beyond. All important language exams in Europe (Cambridge, IETS, TOEFL, etc.) are mapped on
the CEFR.
Although both EAL and CEFR have been developed for contexts in Western countries, they are
instructive because they use competency-based indicators; in particular, the CEFR provides a
framework for multilingual settings. All languages in Europe can be taught, learned, and assessed
using this framework. A similar framework might be helpful for multilingual Africa.

7
 Additive Language Learning for Multilingual Settings

The CEFR indicates a 'threshold' level at which a language student has developed sufficient
competencies to be an independent user of the language. (See Table 1.)
Table 1. Six Levels of the CEFR
A Basic User B Independent User C Proficient User
A1 Breakthrough or beginner B1 Threshold or intermediate C1 Effective operational
proficiency or advanced
A2 Way stage or elementary B2 Vantage or upper
intermediate C2 Mastery or proficiency

The minimal level necessary to function independently in the new language and learn through it is the
B1 threshold or intermediate level. This level is important for bi/multilingual education as it can be
expected that language users have sufficient knowledge of the additional language for it to be used as
a medium of instruction. According to the CEFR, on the global scale the language user:
• “Can understand the main points of clear standard input on familiar matters regularly
encountered in work, school, leisure, etc.;
• Can deal with most situations likely to arise while travelling in an area where the language is
spoken;
• Can produce simple connected text on topics that are familiar or of personal interest;
• Can describe experiences and events, dreams, hopes and ambitions and briefly give reasons
and explanations for opinions and plans” (Council of Europe, n.d., p. 24).
These indicators suggest what content needs to be taught for students to reach this level. It is more
than just vocabulary. These indicators also provide guidance for pragmatic applications—actual use in
real life contexts. However, vocabulary is still a useful indicator. Evidence from various studies show
that vocabulary size can be related to the CEFR levels with some confidence. Because linguistic
variables such as word length and morphology also play an important role in vocabulary size, some
studies have done cross linguistic comparisons. The vocabulary size for the B1 threshold level for
English is about 3,000 words, for French about 2,200, while for Greek about 3,450 (Milton, 2001).
The CEFR was not developed with very young students in mind, however. Instead, it focuses
primarily on lower secondary age and older students (Little 2007). Research that focused on young
students found that these scales do not always reflect how young students learn. Young students
learn language in a “more erratic and recursive” way than is described by the CEFR levels (Enever,
2011, p. 33, 34).
Despite this limitation, a competency-based scale like the CEFR would be helpful for bi/multilingual
education programs as it would allow educators to state what students need to be able to do
(knowledge and skills). This, in turn, would help teachers select content to be taught: specific
domains of vocabulary and types of text for reading and writing. Rather than just making a change in
language of instruction after a certain fixed year, this would allow the new language to be added as a
medium of instruction when the students were ready for it. Depending on contextual issues, this
could be after 4 to 8 years of education.

8
 Additive Language Learning for Multilingual Settings

CASE STUDIES FROM DIFFERENT COUNTRIES

In this section, case studies are presented to further understand what happened in different contexts
when transition to another medium of instruction took place. The review is guided by the following
questions:
• What is the language policy in the country?
• How many years of additive language teaching did students have before the transition took
place?
• What is known about the language competencies the students had when the transition took
place?
• What happened to the student's learning results following transition? What does that reveal
about the competencies the students had developed in the added language?
• Did the students reach the threshold that would enable them to cognitively benefit from
education when the new language is used as a medium of instruction?
ETHIOPIA
The language policy for education in Ethiopia states that children should start education in their
home language. From year 1 onward they learn English as a subject, and from year 3 onward,
Amharic is taught as an additional language. After 4 years of education, the different regions have the
choice of continuing to use the home language or using English as a medium of instruction. The 11
regions in Ethiopia have made different choices. Some have the home language as a medium of
instruction for 4 years of education, others for 6 years of education, and still others have decided to
have the home language as the medium of instruction for 8 years of primary education.
The earliest transition to English as MOI is in grade 5. An analysis of the English language curriculum
for grades 1–4 shows that after 4 years of learning English as a subject, the students will have had to
learn about 400 words (Smith, Stone, & Comings, 2012). Comparing this number to what is known
about the approximate number of words a student needs to know as a threshold for being able to
cognitively benefit from English as a medium of instruction, namely about 3,000 words (Nation,
2006), one can predict that transition at this point to English as MOI will be detrimental to learning
results. The English EGRA results from Ethiopia show that after 4 years of education only 19 percent
of the students in grade 4 (the sum of proficient and advanced level) have reached the required
knowledge and skills of the minimal learning competencies for that grade. (See Figure 2.)
Further analyses of the data show that only 32 percent of the students in grade 4 had reached the
grade-level proficiency of being able to “understand and use a total of about 400 words on different
topics such as everyday objects, food, weather, animals, occupations, buildings, places, parts of the
house, health, safety, tools, parts of the body, and everyday activities”(American Institute for
Research, 2012, p. 22).
These data demonstrate that, based on what is known to be the threshold for the English language,
the learning competencies for grade 4 stated in the current curriculum in Ethiopia are not sufficient.
Moreover, even at the lower level set by the current curriculum, only 19 percent of students reach
the national minimal learning competencies. Therefore, 4 years of teaching English as a subject in
Ethiopia's lower primary schools is not sufficient to reach the curriculum goals of the country or to
prepare young students to have English as an MOI. Other research shows that grade 8 students who
had English as MOI in upper primary school are outperformed in all subjects by students who had
their home language as an MOI and English as a subject (Heugh et al., 2007). Unfortunately, there are
no data yet to show how the students who had their home languages as MOI are doing in secondary
education where the language of instruction is supposed to be English. Further, we do not yet know
if 8 years of learning English as a subject enables Ethiopian students to reach the threshold where

9
 Additive Language Learning for Multilingual Settings

English can be used successfully as a language of instruction. Research might even show that more
years of English as a subject and the mother tongue as an MOI would be better for students.
Figure 1. Data from Ethiopia: EGRA Results (American Institute for Research, 2012)

Grade 2 Grade 3 Grade 4

80%
70%
70% 67%66%
PERCENTAGE OF STUDENTS

60%

50%

40%

30%

20% 13%15%11%
9% 11%13%12%
10% 6% 7%

0%
below basic basic proficient advanced

SOUTH AFRICA
Some very helpful information comes from South Africa, where data have long been collected on
students leaving school at the end of 12 years of education. From 1955 until the early 1970s,
children received their first 8 years of education in their mother tongue while they learned two
other languages (Afrikaans and English) as subjects. After apartheid ended, the language policy was
changed, and children received 4 years of education in their mother tongue/home language and
during the fifth year English became the MOI.
Table 2. South Africa: Pass Rates at Matriculation Level (adapted from Heugh, 2012)

African Percent Overall total

language (%) number of
speaking pass candidates, plus %
Year students rate pass rate

1955 595 43.5

1976 9595 83.7
1979 14574 73.5 85,276 (87%)
1980 29973 53.2 109,897 (75%
1982 70241 48.4 139,488 (69%)
1992 342038 44 448,491 (56%)
1994 392434 49 495,408 (58%)
1997 559,233 (47.4%)
1998 552,862 (49%)

10
 Additive Language Learning for Multilingual Settings

The pass rate on the final school-leaving examinations, which are in English, seems to show that the
change in language policy had an effect on the learning results. The students who had 8 years of
primary school in their home language scored well on the matriculation test, which is in English; this
seems to indicate that they reached the threshold level and could do well in education with English
as the medium of instruction. After the change in policy, however, there was a rapid decline in the
percentage of students who passed the end of school test. South African research shows that the
early switch to English does not work in the majority of South African schools (Heugh, 2012). Other
possibilities are that more changes took place and the instruction was too poor for early transition.
However, at this early transition, “[t]here is [a] gap between students’ English proficiency and the
linguistic demands of learning through the medium of English” (Probyn, 2006, p. 393).
A more recent analysis of data from standardized tests in English and mathematics conducted as part
of the Annual National Assessment over a period of 5 years shows that “after controlling for school-
fixed effects, receiving mother tongue instruction (rather than English instruction) in grades 1, 2 and
3 leads to better English proficiency in grades 4, 5 and 6.”(Taylor & Coetzee, 2013 p. 19). Taylor and
Coetzee show that school-fixed effects, such as school performance, also influence learning
outcomes. In the high performing schools, students learn English better and can therefore start to
use English as MOI earlier than in lower performing schools. This shows again that transition to an
additional MOI should not be expressed in years of education, but rather should be based on
competencies that can be acquired in a certain educational and linguistic context.
CAMEROON
In Cameroon, the policy states that during the first 3 years of education the mother tongue can be
used as MOI, and that in year 4 either French or English are used as MOI. That means that students
have 3 years to gain sufficient knowledge of English or French to be able to benefit from education in
year 4 of primary school.
Interesting data have been presented by the Kom Experimental Mother Tongue Education Pilot
Project (KEPP). This project collected data from students who started education with English as
their MOI and from students who had their mother tongue as their MOI for years 1–3 and learned
English as a subject. This longitudinal study collected data on students from primary school year 1
through year 5 and, thus, was able to see how the students dealt with the transition to a new MOI.
Initial results from KEPP showed that students who had their mother tongue as MOI outperformed
students who had English as MOI in all subjects (Walter & Chuo, 2012; Walter & Trammell, 2010).
Year 4 was the transition year when English became the MOI for all students. The data showed that
having 3 years education with the mother tongue as MOI was not sufficient to prepare the students
for the transition. The students’ scores dropped considerably in years 4 and 5. Interestingly, the
students who had been in an English-only class for all years of education scored lower and seemed
even less prepared for the linguistic demands in class in years 4 and 5 than the KEPP students.
Walter and Chuo (2012) attributed the higher scores of the KEPP students to the fact that their
foundation in education was laid in their mother tongue, as this was the only difference between the
two student populations.
A vocabulary test given to students in regular and KEPP schools in year 5 showed that the students
in the comparison school scored only about 25 percent while the students in the KEPP schools
scored about 33 percent on the test. The scores of both groups are low, but the KEPP students’
scores were comparatively higher. These low scores, however, mean that the majority of students in
both types of schools have not mastered the vocabulary necessary to cope with the academic
content of year 5.
This study shows that 3 years of education with the mother tongue as the MOI and English as a
subject provided better results in English than education with English as the MOI from the beginning.
Thus, a more effective way to learn English is to have the mother tongue as MOI for all subjects.
This seems consistent with the linguistic interdependence hypothesis: what is learned in one
language is transferred to the other language. Further, this study shows that in this context, 3 years

11
 Additive Language Learning for Multilingual Settings

of English as a subject is not sufficient to reach the threshold level to have English added as a MOI.
The students were not ready for the linguistic, cognitive, and academic challenges associated with
English as MOI.
These three case studies show that early transition to a new MOI is not beneficial for education
outcomes in these contexts. In the case of Ethiopia, it was possible to compare the amount of
vocabulary in the curriculum to a known threshold for vocabulary size to enable use of a language as
MOI. The study from South Africa indicates that school factors play a role in the readiness to use a
language as MOI, and policies should consider the individual school and/or linguistic setting.

12
 Additive Language Learning for Multilingual Settings

MULTILINGUAL CLASSROOMS
The case studies above illustrate the advantages of instruction in a known language first for an
extended period of time before transitioning to an additional language as a medium of instruction.
However, in certain situations, more than one mother tongue a may be spoken in a classroom. This
is often the case in urban settings or in areas where two, three or more language communities live
closely together. Several questions arise in these situations: How should language and languages be
treated in classrooms? What would be the best language for instruction?
Different solutions have been applied in different situations. In Ethiopia for example, there are often
two streams of one grade, one stream provides education in the mother tongue, while the other
one uses Amharic as a language of instruction (Gemechu, 2010). This is a practical solution when
there is one dominant mother tongue and there is a group of students who do not speak the
mother tongue being taught, or who share another mother tongue. In other situations, where there
are a variety of mother tongues and it might not be possible to provide individualized mother tongue
instruction for each student, research is needed on methods for determining the most effective
language of instruction. It might be that most students are bi- or tri lingual and that all languages
would help students cognitively benefit from education. However, it could also be that the young
students only speak their mother tongue. In that case, there are a few options to choose from for
instruction: use the languages (of the larger group of students) as a medium of instruction; use an
African language (language of wider communication) that is not spoken by any of the students; use a
foreign language that is later on used in school as a medium of instruction (English, French or
Portuguese). The first option is often criticized because there is a belief that students whose
language is not chosen will be disadvantaged. The second option is often criticized as well; it is
argued that it would be better to use one of the European languages, as they will be used anyway
and it gives students a head start with that language.
Research from Africa is starting to give insight on what is best for students when more than one
mother tongue is spoken. In Cameroon, a small group of Fulfulde-speaking people are living in an
area where most people speak Kom. For the Kom community, mother tongue education had been
developed and it would have been easiest for the Fulfulde students to start in classrooms where
Fulfulde would be the MOI. However, as the population of Fulfulde speakers was small, the
education officials decided using the language as MOI would not be feasible. The Fulfulde students
had two options: joining an English medium school or a Kom medium school. Research data showed
that the Fulfulde-speaking students in the Kom medium schools outperformed the Fulfulde-speaking
students in the English medium school by 72 percent in Grade 2, and by 127 percent in grade 3
(Walter, n.d.). Thus, the assumption that an international language as the MOI in mixed language
classes does not hold in this case.
Other research shows that when there is a large group of mother tongue speakers of one language,
children living within that context benefit from mother tongue-based multilingual education. In an
environment with mixed language groups, using a language of wider communication is beneficial for
children (Walter, n.d.).
This research indicates that in multilingual environments, it is important to understand the context
to determine what type of education would be most practical and beneficial for children in that
situation. When two language communities are living together it might be possible to have two
streams in one grade, each providing education through the mother tongue that applies to children
in the group. It may be that children in these contexts are bilingual and know, or are familiar with,
the language of ‘the other group’. Thus, they can also benefit from education in the ‘other group’
language. A European language should only be considered as a MOI when the aforementioned
options are not possible.

13
 Additive Language Learning for Multilingual Settings

SUMMARY
This paper reviews existing research and provides case studies in three countries to investigate
approaches to providing bilingual and multilingual education. Preliminary answers to the following
questions are explored: When can the MOI successfully include other languages besides mother
tongue? How can this threshold level be determined? Is this threshold different for different
languages?
Two measures for threshold levels have been highlighted: one based on vocabulary size, the other
based on competencies. Vocabulary size as a measure of language learning is helpful for providing
guidance on how many word families should be learned in order to reach the threshold, which
differs across languages. To efficiently determine the words students should learn, focus on words
used most frequently in that language. The selection of words, therefore, depends on the language.
Further, vocabulary knowledge (and size) is a good indicator for reading comprehension. Unassisted
comprehension can only take place when at least 95 percent of the words in text are known. It is
crucial that students acquire sufficient vocabulary before a language is used as a MOI. Research
suggests that in school settings, students learn between 300–600 words a year.
However, language learning and language use depends on more than just vocabulary. Understanding
the structure of a new language and how text is built in the language are important factors that aid
transfer of reading skills from one language to another. Vocabulary size does, however, provide an
indication of when a student has reached the threshold of being able to function and learn through
the new language. Information on competencies, such as the standards in the Common European
Framework Reference for Languages (CEFR), can help curriculum developers by setting objectives to
guide content decisions for curricula, textbooks, and other materials. The CEFR levels can also be
used to guide decisions on vocabulary size.
There is no single solution for language teaching. Rather, several contextual factors need to be
considered when thinking about multilingual education. Evidence shows that vocabulary size as a
measure is language dependent. Thus, for each language that is taught, it is necessary to establish
what the vocabulary threshold level is and to identify the most frequent words and/or morphemes
that students should learn.
As the case studies and literature show, other factors also influence how successfully students learn
a new language. The linguistic distance between the mother tongue of the students and the new
language determines the ease and pace with which the new language can be learned. The greater the
linguistic distance, the longer it takes to learn the new language. Also, the number of contact hours,
the language level of the teachers, the use of the new language outside the classroom, and the
perceived usefulness and value of the languages all play important roles in how quickly and effectively
students can learn a new language. Young students learn at different speeds in different contexts.
Therefore, the importance of each of these variables differs across contexts; different outcomes can
be expected in terms of how long a student needs to be taught a language before it can become a
MOI across settings.
Research suggests that it is not effective to teach a new language as a subject for a few years and
then switch to that language as MOI without first establishing that students have reached a critical
threshold level. Case study results have shown that the drop in learning outcomes is considerable
otherwise. Unfortunately, no one method or approach will work for all contexts, so educators must
be flexible and understand that different strategies available. Research has also shown that students
can acquire the critical threshold level and learn a language well without it being used as a MOI in
education.
In mixed language situations, it is important to understand the context (i.e., level is bilingualism,
language distance) to target policy suggestions towards providing the most beneficial education for
children. The use of an additional language as MOI cannot be expressed in years of education; the
educational and linguistic context in which the additional language is learned needs to be considered.

14
 Additive Language Learning for Multilingual Settings

RECOMMENDATIONS
Interest in multilingual education is high throughout the world. For example, the United Nations
Economic, Scientific and Cultural Organization (UNESCO) is developing various initiatives to
promote education in mother tongue and sponsored International Mother Language Day in February
2014. 3 There is a need for additional practical information and ideas that can be used for planning
and structuring multilingual education programs.
Before starting a multilingual education program, it is essential to evaluate the approaches and
determine which is best suited to learning an additional language and using that language as a MOI in
the specific context. The issue is complex, but the factors that influence language learning have been
identified and can be strategically addressed to benefit education, rather than hindering it; as is the
case in many contexts. The recommendations for multilingual education presented in this report
seek to contribute to the body of knowledge that can be adapted and implemented in different
country contexts.
The following are recommended as preliminary guidelines for effective multilingual education:
Time
• Give young students time to learn to read in a language they know well.
• Give young students time to learn sufficient vocabulary in the new language, then have them
learn to read in the new language, making use of their knowledge of reading in their first
language.
Vocabulary and competencies
• Develop a competency-based scale that states what students need to be able to do
(knowledge and skills) in the new language at different stages of their education.
• Determine the threshold level of vocabulary and competencies that enables students to
benefit from education in each second or foreign language (e.g., English, French, Portuguese,
Wolof, Kiswahili, Amharic, Hausa) to be used as a MOI in school.
Context
• Understand the context in which the new language is to be taught and consider this for
planning when students are expected to be able to understand the new language as an MOI.
• At the national level, adopt a flexible policy that guides schools in choosing when to
transition their students by adding the new language as MOI. Language tests based on
competencies will be helpful in determining when young students are ready to transition.
Materials
• Ensure that the reading methodologies used for the mother tongue and the added language
complement each other and take into account the linguistic variables of the languages used in
the bi/multilingual education setting.
• Ensure that the language teaching methodology takes into the account the similarities and
differences between the students' mother tongue and the added language(s), at least for the
first few years.

3. For more information on UNESCO's work on multilingual education, visit

http://www.unesco.org/new/en/education/themes/strengthening-education-systems/languages-in-education/multilingual-education/

15
 Additive Language Learning for Multilingual Settings

Table 3. Factors that influence the pace of language acquisition

Factor Reduced pace of Faster pace of

acquisition/longer period acquisition/shorter
of language learning period of language
learning

Time in a week less than 3 periods a week more than 3 periods a

week
Vocabulary size necessary for a larger vocabulary size is a smaller vocabulary
threshold level (language dependent) necessary for threshold level size is necessary for
threshold level
Competences necessary to function the context makes it the context allows for
at threshold level necessary for transition to transition to take
take place at a higher grade place at a lower grade
Linguistic distance between the L1 linguistic distance is large; a linguistic distance is
and the additional language longer period of language less or small; a
learning is necessary shorter period of
language learning is
sufficient
Writing systems of the L1 and the writing systems differ greatly writing systems differ
additional language(s) (orthographic depth, syllable slightly (orthographic
structure, etc.) depth, syllable
structure, etc.)
Script of the L1 and the additional the scripts differ greatly there is no or a small
language(s) difference between
the scripts
Language level of teacher that language level is below B2 language level is B2 or
teaches the additional language C1 or higher
The number of students in class more than 15 students less than 15 students
The interaction in the additional few interactions results in a many interactions
language classroom slower pace for language increase the language
acquisition acquisition pace
The starting age of the students very young (between 4-10 about 10+ years of
years of age) age
The perceived value of the additional the perceived value of the the perceived value of
language additional language is low; it the additional language
takes longer to learn it well is high; enhancing
acquisition pace
The additional language is a second when the additional language when the additional
or foreign language is foreign, language language is a second
acquisition takes longer language, pace of
acquisition is
enhanced

16
 Additive Language Learning for Multilingual Settings

• For policy makers it is important to understand factors that influence the acquisition of an
additional language. Policy makers can use this information to assess the potential success of
an existing language policy and curriculum and adjust it so that success is enabled for all
children who enter school. The table above summarizes the key factors discussed in this
paper and how each influences language acquisition.
• There are many factors that influence language learning, and these factors affect each other
as well. The context plays a role in the degree in to which these factors influence language
acquisition. Therefore, it is important to conduct research and understand the context in
which a policy and curriculum are being implemented.
For many additional languages, we do not yet know the most frequent words that should be taught.
Also, a better understanding is needed of how differences between writing systems and scripts
influence language acquisition in different contexts. Finally, more information is needed on what
students acquire after each of education in different contexts, in terms of amount of vocabulary and
different competencies. This would provide better guidance to policy makers on when students are
ready to receive education through an additional language.
Based on evidence from the field, current language policies can be adapted to show respect for the
linguistic backgrounds and heritages of people in different language groups. Evidence also can provide
policymakers with clearer information on which sectors of the education systems in their country
contexts need to be changed to ensure the best approach to language teaching.

17
 Additive Language Learning for Multilingual Settings

BIBLIOGRAPHY
Alidou, H., Boly, A., Brock-utne, B., & Satina, Y. (2006). Optimizing learning and education in Africa –
the language factor, 1–186.
Almabruk, A. A. A., Paterson, K. B., McGowan, V., & Jordan, T. R. (2011). Evaluating effects of
divided hemispheric processing on word recognition in foveal and extrafoveal displays: the
evidence from Arabic. PloS One, 6(4), e18131. doi:10.1371/journal.pone.0018131
American Institute for Research. (2012). Ethiopia English Early Grade Reading Assessment : Data Analytic
Report. Retrieved from http://pdf.usaid.gov/pdf_docs/pnady822.pdf
Arab-Moghaddam, N., & Sénéchal, M. (2001). Orthographic and phonological processing skills in
reading and spelling in Persian/English bilinguals. International Journal of Behavioral Development,
25(2), 140–147.
Asfaha, Y. M., Kurvers, J., & Kroon, S. (2009). Grain size in script and teaching: Literacy acquisition in
Ge’ez and Latin. Applied Psycholinguistics, 30(04), 709. doi:10.1017/S0142716409990087
Bossers, B. (1991). On thresholds, ceilings and short-circuits: the relation between L1 reading, L2
reading, and L2 knowledge. Aila Review, 8, 45–60.
Bot, K. De, & Herder, A. (2008). Effecten van vroeg. Taal Lezen Primair, (1), 2–3.
Carrell, P. L., Devine, J., & Eskey, D. E. (Eds.). (1998). Interactive approaches to second language reading.
Cambridge: Cambridge University Press.
Chenoz, J. (2008). Achievements and challenges in bilingual and multilingual education in the Basque
Country. Aila Review, 21, 13–30.
Chiswick, B. R., & Miller, P. W. (2004). Linguistic distance : A quantitative measure of the distance
between English and other languages. Bonn. Retrieved from http://ftp.iza.org/dp1246.pdf
Chiswick, B. R., & Miller, P. W. (2008). A test of the critical period hypothesis for language learning1.
Journal of Multilingual and Multicultural Development, 29(1), 16–29. doi:10.2167/jmmd555.0
Coderre, E. L., Filippi, C. G., Newhouse, P. A., & Dumas, J. A. (2008). The Stroop Effect in Kana and
Kanji Scripts in Native Japanese Speakers: An fMRI Study. Brain Language, 107(2), 124–132.
doi:10.1016/j.bandl.2008.01.011.The
Cook, V. (2003). Effects of the second language on the first. Clevedon: Multilingual Matters.
Coulmas, F. (1996). The blackwell encyclopedia of writing systems. BlackWell Publishing.
Council of Europe. (n.d.). 1 The Common European Framework in its political. Retrieved from
http://www.coe.int/t/dg4/linguistic/Source/Framework_en.pdf
Cummins, J. (1979). Linguistic interdependence and the educational development of bilingual
children. Review of Educational Research, 49(2), 222–251. doi:10.3102/00346543049002222
Djigunovich, J. M., & Vilke, M. (2000). Eight years after; wishful thinking or the facts of life. In J. Moon
& M. Nikolov (Eds.), Research into teaching English to young learners: International perspectives (pp.
67–86). Pécs: University Press Pécs.
Djité, P. G. (2008). The Sociolinguistics of Development in Africa. Clevedon: Multilingual Matters.
Duibhir, P. O., & Cummins, J. (2012). Towards an integrated language curriculum in early childhood and
primary education (3-12 years).
Education First English Proficiency Index. (2013). Retrieved from
http://www.ef.com/__/~/media/efcom/epi/2014/full-reports/ef-epi-2013-report-master-new.pdf

18
 Additive Language Learning for Multilingual Settings

Enever, J. (Ed. . (2011). Early language learning in Europe.

Enever, J., Damhuis, R., Lindgren, E., Lopriore, L., Lundberg, G., Djigunovic, J. M., … Mestres, E. T.
(2008). Tussentijdse resultaten 2007‐2008 (pp. 1–4).
Enever, J., Moon, J., & Raman, U. (2009). Young learner english language policy and implementation:
International perspectives. Reading: Garnet Publishing Ltd.
Eyckmans, J. (2004). Vocabulary size. Measuring receptive vocabulary size reliability and validity of the
yes/no vocabulary test for French-speaking Learners of Dutch. Katholiek Universiteit Nijmegen.
Retrieved from http://www.lotpublications.nl/publish/articles/000810/bookpart.pdf
Fontecha, A. F. (2009). Spanish CLIL: Research and official actions. In Y. Ruiz de Zarobe & R. M.
Jimenez Catalan (Eds.), Content and Language Intergrated Learning; Evidence from Research in
Europe. (pp. 3–21). Bristol, UK.: Multilingual Matters.
Gemechu, D. T. (2010). The implementation of a multilingual education policy in Ethiopia: The case of
Afaan Oromoo in Primary Schools of Oromia Regional State. University of JYVÄSKYLÄ.
Gorter, D., & van der Meer, C. (2008). Developments in bilingual Frisian-Dutch education in
Friesland. Aila Review, 21.
Heugh, K. (2012). The Case Against Bilingual and Multilingual Education in South Africa. PRAESA,
Occasional Papers. Cape Town: PRAESA. Retrieved from http://www.praesa.org.za/occasional-
paper-6-the-case-against-bilingual-and-multilingual-education-in-south-africa/#.Ug5dmW37_AA
Heugh, K., Benson, C., Berhanu, B., & Yohannes, Mekonnen, A. G. (2007). Final report study on
medium of instruction in primary schools in Ethiopia.
Hirsh, D., & Nation, P. (1992). What vocabulary size is needed to read unsimplified texts for
pleasure? Reading in a Foreign Language, 8(2), 689–696.
Johnstone, R. (2002). Addressing ’ the age factor ': Some implications for languages policy.
Strasbourg: Council of Europe.
Kieffer, M. J., & Lesaux, N. K. (2007). The role of derivational morphology in the reading
comprehension of Spanish-speaking English language learners. Reading and Writing, 21(8), 783–
804. doi:10.1007/s11145-007-9092-8
Language Policy Unit. (n.d.). Common European framework of reference for languages: Learning ,
teaching , assessment. Strasbourg. Retrieved from www.coe.int/lang-CEFR
Little, D., Goullier, F., & Hughes, G. (2011). The European Language Portfolio: The story so far
(1991–2011). STRASBOURG: Council of Europe. Retrieved from
http://www.coe.int/t/dg4/education/elp/elp-
reg/Source/Publications/ELP_StorySoFar_July2011_Final_EN.pdf
Milton, J. (2001). The development of vocabulary breadth across the CEFR levels. Eurosla Monographs
Series 1 Communicative Proficiency and Linguistic Development, (1), 211–232. Retrieved from
http://eurosla.org/monographs/EM01/211-232Milton.pdf
Moreno, F. X. V. i. (2008). Language in Education policies in the Catalan language area. Aila Review,
21, 31–48.
Mothibeli, A. (2005, October). Cross-Country Achievement Results from the SACMEQ 11 Project –
2000 to 2002. A Quantitative Analysis of Education Systems in Southern and Eastern Africa.
Edusource Data News. No. 49.
Muñoz, C. (2008). Age-related differences in foreign language learning. Revisiting the empirical
evidence. IRAL - International Review of Applied Linguistics in Language Teaching, 46(3), 197–220.

19
 Additive Language Learning for Multilingual Settings

Nag, S. (2007). Early reading in Kannada: The pace of acquisition of orthographic knowledge and
phonemic awareness. Journal of Research in Reading, 30, 7–22.
Nation, I. S. P. (2006). How Large a Vocabulary Is Needed For Reading and Listening ? The Canadian
Modern Language Review, 63(1), 59–82.
Nation, I. S. P., & Wang, M. (1999). Graded readers and vocabulary. Reading in a Foreign Language.,
12, 355–380.
Nikolov, M. (2000). Issues in research into early FL programmes. In J. Moon & M. Nikolov (Eds.),
Research into teaching English to young learners: International perspectives (pp. 11–48). Pécs:
University Press Pécs.
Nikolov, M. (2006). Relationships Between Language Achievements in English and German and
Classroom-related Variables, 197–224.
Nikolov, M. (Ed.). (2009). The Age Factor and Early Language Learning. The Age Factor and Early
Language Learning. Berlin: Mouton de Gruyter.
Orosz, A. (2009). Growth of young learners’english vocabulary size. In M. Nikolov (Ed.), Early
Learning of Modern Foreign Languages, Processes and Outcomes (pp. 181–194). Bristol, UK.:
Multilingual Matters.
Ouane, A., & Glanz, C. (2006). Mother tongue literacy in Sub-Saharan Africa, (June), 1–9.
Piper, B. (2010). Kenya Early Grade Reading Assessment Findings Report July 2010, (July).
Piper, B., & Van Ginkel, A. J. (2014). Reading the Script: Ethiopian language scripts’ impact on letter
and word identification. Under Submission.
Probyn, M. (2006). Language and Learning Science in South Africa. Language and Education, 20(5),
391–414. doi:10.2167/le554.0
Qualifications and Curriculum Authority. (2000). 2000 A language in common : Assessing English as
an additional language. London: Qualifications and Curriculum Authority 2000.
Seymour, P. H. K., Aro, M., & Erskine, J. M. (2003). Foundation literacy acquisition in European
orthographies. British Journal of Psychology (London, England : 1953), 94(Pt 2), 143–74.
doi:10.1348/000712603321661859
Smith, C., Stone, R., & Comings, J. (2012). Literacy Policy and Practice the TELL Program and,
(March).
Taylor, S., & Coetzee, M. (2013). Estimating the impact of language of instruction in South African
primary schools : A fixed e ff ects approach. Stellenbosch.
Thijs, A., Trimbos, B., Tuin, D., Bodde, M., & de Graaff, R. (2011). Engels in het basisonderwijs.
Enschede: SLO. Retrieved from http://engelsbo.wordpress.com/2012/02/04/vakdossier-engels-
in-het-basisonderwijs/
Tragant, E. (2010). Multilingualism in Europe : on the political agenda.
Unsworth, S., Bot, K. De, Persson, L., & Prins, T. (2012). Eindverslag FLiPP November 2012,
(November).
Uwezo. (2010). Are our children learning ? Annual learning assessment, Kenya 2010; Summary and Key
Findings. Nairobi.
Van Ginkel, A. J. (2008). Towards a methodology of transfer reading. Leeds Metropolitan University.
Walter, C. (2004). Transfer of Reading Comprehension Skills to L2 is Linked to Mental
Representations of Text and to L2 working memory. Applied Linguistics, 25, 315–339.

20
 Additive Language Learning for Multilingual Settings

Walter, S. L. (n.d.). Mother Tongue-based Education in Developing Countries : Some emerging

insights. Unpublished research report. Available from author. Retrieved from
http://www.globalpartnership.org/media/library/blog/Steve-Walter-Mother-TongueInsights.pdf
Walter, S. L., & Chuo, K. G. (2012). The Kom Experimental Mother Tongue Education Pilot Project Report
for 2012 (pp. 2–12). Retrieved from http://www.mlenetwork.org/sites/default/files/The Kom
MLE Project 2012.pdf
Walter, S. L., & Trammell, K. (2010). The Kom Experimental Mother Tongue Education Project
Report for 2010 Summary test results from 2010, 1–22.
Ziegler, J. C., Bertrand, D., Tóth, D., Csépe, V., Reis, A., Faísca, L., … Blomert, L. (2010).
Orthographic depth and its impact on universal predictors of reading: a cross-language
investigation. Psychological Science, 21(4), 551–9. doi:10.1177/0956797610363406

21