This article was downloaded by: [University of Auckland Library]

On: 09 April 2015, At: 16:46

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number:
1072954 Registered office: Mortimer House, 37-41 Mortimer Street,
London W1T 3JH, UK

Journal of Multilingual and

Multicultural Development
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/rmmm20

Issues in Ethiopian
Language Policy and
Education
Thomas Bloor & Wondwosen Tamrat
Published online: 29 Mar 2010.

To cite this article: Thomas Bloor & Wondwosen Tamrat (1996) Issues
in Ethiopian Language Policy and Education, Journal of Multilingual and
Multicultural Development, 17:5, 321-338, DOI: 10.1080/01434639608666286

To link to this article: http://dx.doi.org/10.1080/01434639608666286

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all
the information (the “Content”) contained in the publications on our
platform. However, Taylor & Francis, our agents, and our licensors
make no representations or warranties whatsoever as to the accuracy,
completeness, or suitability for any purpose of the Content. Any
opinions and views expressed in this publication are the opinions and
views of the authors, and are not the views of or endorsed by Taylor
& Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information.
Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities
whatsoever or howsoever caused arising directly or indirectly in
connection with, in relation to or arising out of the use of the Content.
 This article may be used for research, teaching, and private study
purposes. Any substantial or systematic reproduction, redistribution,
reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access
and use can be found at http://www.tandfonline.com/page/terms-and-
conditions
Downloaded by [University of Auckland Library] at 16:46 09 April 2015
 Issues in Ethiopian Language Policy and
Education
Thomas Bloor
Lang uage Studies Unit, Aston University, Birm ingham B4 7ET, UK
Wondwosen Tamrat
Kotobe College of Teacher Education, Addis Ababa, Ethiopia
In Ethiopia, the overthrow of the military government of the Dergue in 1991 ushered
in a period of political reforms which are still in progress, including a policy of
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

recognition of a range of indigenous languages other than the traditionally dominant

national language, Amharic. This paper addresses the question of language planning
and education in this pluralist, multilingual state. It presents a profile of some of the
many languages in use within the country, both indigenous and foreign, and briefly
considers their historical role in the education system, touching on the associations of
Amharic and the classical Ethiopian language Giiz with Orthodox ‘Coptic’ Christian-
ity, and of Arabic with Islam. An evaluative appraisal of the new policy is offered
which, whilst acknowledging the enlightened nature of the reforms in light of current
political and linguistic concerns for ‘linguistic human rights’ (Phillipson & Skutnabb-
Kangas, 1995), considers a number of major obstacles to the implementation of the
policy, and questions in particular the moves towards the preference of the Roman
alphabet to the well-established Ethiopic syllabic writing system. The article points
out that, contrary to the intentions of the reforms, the outcome could benefit English
at the expense of a significant African language.

Introduction
Recent political changes in Ethiopia have brought language planning into
prominence and policies are being forged which have significant implications for
the future of the nation. The most central changes affect the relative status of
Amharic (the official national language), the many other indigenous languages,
and English.
It has been argued that the widespread perception of a ‘multiplicity of
languages as a “problem” may only be an illusion’ (Khubchandani, 1984: 102),
and it is true that in terms of interaction in buying and selling, practical work,
transport, medical treatment and other social activities, pluralistic societies
operate with a remarkable flexibility and efficiency, determining such issues as
the choice of lingua franca with pragmatic ease. People with little or no formal
education manipulate two or more languages without much apparent difficulty
and the work gets done.
However, it cannot be denied that pluralism presents massive difficulties for
policy-makers and raises many questions to which there are no easy answers.
The contrasting demands of instrumental and sentimental criteria are sometimes
incompatible. We certainly do not foresee any easy solution to the complex
language problems of Ethiopia.
It is a truism that language planning cannot be considered independently of
the sociolinguistic setting in which it operates; we shall therefore devote some
0143-4632/96/05 0321-18 $10.00/0 © 1996 T. Bloor & W. Tamrat
JOURNAL OF MULTILINGUAL AND MULTICULTURAL DEVELOPMENT Vol. 17, No. 5, 1996
321
 322 Journal of M ultiling ual and M ulticultural D evelop m ent
space to the description of the linguistic history and geography of Ethiopia, and
comment on relevant socio-political features. This is especially necessary here
because, although many of her language problems are typical of African nations
and indeed of pluralistic societies in general, Ethiopia is in several important
respects interestingly different from most other African countries and cannot be
dealt with on the basis of presuppositions derived from observing her neigh-
bours.

Population and Languages

Ethiopia (formerly known as Abyssinia) is historically and culturally unique
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

in Africa in at least two important respects. Firstly, although not untouched by

European colonialism, it is the only African country that was not colonised by a
European power during the nineteenth-century scramble for Africa, and,
secondly, it is the only African country where the dominant religion is an
indigenous form of Christianity dating back to the fourth century AD. Both these
factors have played a crucial role in the linguistic profile of the country. Other
unique features possessed by this exceptional land include its writing system,
Ethiopic script.
The estimated population in 1970 was 25 million, although figures have never
been very reliable. In the succeeding 25 years, there have been significant
changes, including, in spite of the well-publicised famines and bloody civil war,
a large increase in population. According to a government estimate (National
Office of Population, 1993), it is now more than doubled at 53.2 million. Addis
Ababa, the capital city, has a population of over 2.2 million if projections reported
eight years ago are correct (National Office of the Population and Housing
Census Commission, 1987). In the census of 1967 it was calculated as 600,000 and,
even at that time, over half the population had migrated to the capital from
outside. In spite of increasing urbanisation, however, only about 15% of the
national population is estimated to be living in towns (National Office of
Population, 1993).
One complication in the discussion of Ethiopia is the status of Eritrea, currently
an independent state after prolonged warfare — first against the government of
Haile Selassie and then, more intensely, against the Dergue, the military junta
which overthrew him. Eritrea has been at various times an Italian colony, a UN
mandated territory under British protection, part of a federation with Ethiopia
and a province of Ethiopia. Demographic figures prior to 1990 normally include
Eritrean data as an integral part of general Ethiopian data.
The most extensive study of languages in Ethiopia was the Ford Foundation
Survey led by Charles Ferguson and published as Bender et al. (1976). It is
unfortunate that nothing comparable has been carried out since then, as there
have been significant changes. It is hardly necessary to point out that there are
major problems in determining whether two given speech communities speak
related languages or dialects of the same language, and that the assignment of
languages to language families is fraught with difficulty (see Edwards, 1994).
Last (1954: 1) speaks of ‘200 languages or variations of languages’ (‘variations’
presumably meaning ‘dialects’). The figure of 70 is often mentioned, but
 Issues in Ethiopian Languag e Policy and Educ ation 323
Wedekind (1994) identifies 99 languages spoken as mother tongues with varying
names totalling around 420 (cf. Edwards, 1994 on multiple naming).
Most of the languages belong to the Afro-Asiatic superfamily. Within this
superfamily, three major language families are represented: Semitic, Cushitic
and Omotic. The family with the largest number of languages is Cushitic, but the
Semitic family could have a larger number of mother-tongue speakers (49.7% of
the total, according to the Ford Foundation Survey, as opposed to 43.5% for
Cushitic and only 5.3% for Omotic.) An entirely different superfamily, Nilo-
Saharan, found mainly near the Sudan border in the extreme west and somewhat
more extensively in the south-west, is represented by several families, each with
a number of languages (such as Anyuak and Nuer) but all spoken by small
populations, probably totalling only 1.5%.
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

Major Indigenous Languages

Ferguson (1967) posits three criteria, any one of which qualifies a language as
a major language in a country: (a) being spoken by over a million speakers or by
25% of the population; (b) being an official language; (c) being the medium of
education of over 50% of secondary school graduates. In 1970, he listed Amharic,
1
Oromo and Tigrinya as meeting criterion (a) (Ferguson, 1970). There are now
five indigenous languages which qualify under criterion (a) (see Table 1). Of
these, Amharic and Tigrinya have substantial written traditions and belong to
the Semitic language family, whereas Oromo is Cushitic, like Somali. Amharic
qualifies additionally on criterion (b) (official language).
Recent official sources speak of more than 90 different ‘ethnic groups’ and
indicate the proportions given in Table 1. Oversimplifying grossly, Tigrinya as a
mother tongue is predominant in Eritrea and northern Ethiopia (Tigray
province), Amharic in the north and centre and Oromo in the centre and south;
Gurage is mainly concentrated in a small area to the south of Addis Ababa;
Somali in the south-east, and Sidama and Welaita in the south; but the situation
is much more complicated than that suggests.
It can be seen that in terms of number of speakers, Oromo and Amharic are
well ahead of the rest. Now that Eritrea is a separate state, Tigrinya has fallen
Table 1 Estimated mother-tongue speakers in Ethiopia (based on National Office
of Population 1993: 1)

Language est. %age millions

Oromo 29.1 15.48
Amharic 28.3 15.06
Tigrinya 9.7 5.16
Gurage 4.5 2.39
Somali 3.8 2.02
Sidama 3.0 1.60
Welaita 2.6 1.38
Others 19.0 10.11
Total 100 53.2
 324 Journal of M ultiling ual and M ulticultural D evelop m ent
further behind numerically in Ethiopia as now defined, but remains third.
Tigrinya is also significant for its level of development in literature, literacy,
standardisation, etc., and is politically very important in that it is the mother
tongue of key elements (formerly members of the Tigray People’s Liberation
Front) in the current leadership of the Ethiopian People’s Revolutionary
Democratic Front.
A recent survey of bilingualism and multilingualism in Addis Ababa (Senayit,
1994) reveals that 73.2% of Amharic speakers in the capital claim to be
monolingual as opposed to 0.6% of Gurage, 0.3% of speakers of various Omotic
languages, 0.2% of Oromo, and no Tigrinya speakers. For bilinguals, Amharic
was the other language of 95.1% who could speak an Omotic language, 92.1%
2
Oromo speakers, 90.5% Gurage speakers and 88.9% Tigrinya speakers, and all
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

multilinguals had Amharic in their repertoire. In most other regions, the figures
could be drastically different of course, but Amharic is very clearly the primary
language of the capital, even though the majority language in the surrounding
countryside is Oromo, the traditional Amhara heartlands being further north.
Amharic has for long had a favoured position in Ethiopia as the language of
government and to a great extent as the non-liturgical spoken language of
religion. Under Haile Selassie and under the Dergue, it was the official medium
of primary education, even in Tigrinya-speaking regions, and it was perceived
by these governments as the language of national unity. It was used in
administration and courts of law throughout the country, often with very
precarious interpreting support for non-Amharic speakers, though judges were
in principle expected to know the local languages. It is featured on postage
stamps, air letters and banknotes, and the majority of newspapers are printed in
it. It is the most widespread lingua franca in the country, particularly in the
towns. The policy, with regard to foreign missionary activity, favoured the
spread of Amharic since it limited such activity to non-Christian areas, which
tend to be non-Amharic-speaking, and insisted on the use of Amharic as the
medium of religious teaching (Cooper, 1976).

English and Other European Languages

English, though not exactly an official language, has for many years had a
special status in Ethiopia, especially since the liberation from Italian occupation
in 1941. In so far as any official publication uses a foreign language, it is likely to
be English, and it appears on currency and postage stamps alongside Amharic.
English meets Ferguson’s criteria for classification as a major language in terms
of criterion (c) (secondary school graduates).
However, in view of the small proportion of the population exposed to
secondary education, it is grossly misleading to list English alongside the major
indigenous languages. Schmied observes that calling African nations ‘Anglo-
phone’ is obviously a great exaggeration since, with the exception of Liberia and
South Africa, English is spoken by only ‘an educated minority’ (1991: 27).
Phillipson (1992) makes a similar point. Presumably, this situation is changing
as education becomes more widespread and more children are exposed to
English (though the question of what standards are attained may be relevant).
 Issues in Ethiopian Languag e Policy and Educ ation 325
But the minority factor is more marked in the case of Ethiopia than in most other
African nations where English is a medium of education.
The main reason is that the role of English outside the education system is
considerably more limited than in most so-called Anglophone countries in
Africa. Schmied’s (1991) analysis of domains of English in East African states
shows Ethiopia to have the smallest number of domains: 5 out of 15 compared
with Zambia, Zimbabwe, Malawi, Uganda and Kenya, each of which has 14. The
five domains where Schmied identifies English being used in Ethiopia are:
secondary school, radio, newspaper, local novels and business correspondence,
and it must be stressed that, except in secondary schools and to some extent
business correspondence, the use of English even in these few domains is
enormously outweighed by that of Amharic. This difference is not surprising,
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

given the colonial history of the countries mentioned.

Another important European language has been Italian, though its signifi-
cance appears to be rapidly diminishing. Italy had colonised Eritrea by 1889 and
only gave up its attempt to expand into the rest of Ethiopia when resoundingly
defeated by an Ethiopian army under Menelik II at the battle of Adwa in 1896,
one of the few notable victories won by African forces against nineteenth-century
European colonisers. Under Mussolini’s fascist dictatorship, Italy invaded
Ethiopia again in 1935 and occupied most of the country until 1941, when the
Italians were defeated by Ethiopian and British forces. Subsequently Eritrea
became part of Ethiopia though some Italians remained, mainly in Eritrea but
some in Ethiopia proper. A form of Italian had developed with some features of
a pidgin but with little lexical or grammatical input from indigenous languages
(Habte-Mariam, 1976).
The other historically significant European language in Ethiopia is French but,
again, its role has greatly diminished over the years. The Emperor Haile Selassie
(1891–1975) received his education in French, like many upper-class Ethiopians
of his generation, and in the 1960s France under de Gaulle invested in attempts
to re-establish the language, though with little success. There is a long-established
French lycée in Addis Ababa. French has also normally been an optional subject
in the state system but it now has limited significance.
The languages of other noticeable, though small, European minorities (Greeks
and Armenians) have had little impact and these people were generally fluent
speakers of indigenous languages. Their presence diminished, as did that of Arab
shopkeepers, during the period of government by the Dergue, when private
businesses were suppressed.

Special Languages: Arabic and Giiz

For Europeans before the eighteenth century, Ethiopia was the semi-mythical
land of Prester John, the priest-king, an island of Christianity fending off Islam,
but modern Ethiopia also has a Muslim population, though proportionally
diminished with the separation of Eritrea. The recent National Report mentions
a figure of 33% Muslim as compared with 54% Ethiopian Orthodox Christian,
5% Protestant, 1% Catholic and 6% ‘some form of traditional African religion’
(National Office of Population, 1993: 1).
Arabic is thus significant as what Ferguson calls a special language, that is, one
 326 Journal of M ultiling ual and M ulticultural D evelop m ent
used for religious purposes, etc. Arabic in Ethiopia and Eritrea is primarily
Koranic, playing a significant role in religious practices, literacy and recitation
rather than as a means of secular social interaction. Where it is taught (in Koranic
schools), it has not traditionally been the medium of instruction (but see below).
There are also some native speakers, mainly traders of Yemeni origin on the coast
and, in the west, immigrants (from the nineteenth century onwards) from the
Sudan. Ferguson speculated in 1970 that these must run into tens of thousands,
and observed that Arabic has a small lingua franca function, in markets and
mainly among Muslims of different mother tongues. It is important to note that
at the time of Ferguson’s comment, Eritrea was deemed to be part of Ethiopia,
and Arabic influence is stronger there.
For Christian Ethiopia, the special language is Giiz. This is the classical
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

Ethio-semitic language of the Axumite empire, now a dead language but

crucially important in the liturgy and literature of the Ethiopian Orthodox
(Coptic) Church and still an important lexical resource for official and technical
terms, including Marxist and other politico-economic vocabulary. There are
striking parallels between Giiz in Ethiopia and Latin in Western Europe or
Sanskrit in India. Some scholars believe it peaked as a literary medium between
the seventh and thirteenth centuries, some time after it had ceased to exist as a
mother tongue. It is written in Ethiopic script (see below).
Traditionally, literacy was acquired in priest schools on a pattern similar to
that of the Koranic schools, that is, by rote and memorisation in a language which
the learner did not speak and was not expected to speak. Literacy in Amharic or
Tigrinya was largely an accidental spin-off of literacy in Giiz. As in medieval
Europe, the church professionally occupies a much larger proportion of the
population than is common in most modern communities, and an individual
whose career is in the church as priest, deacon or wandering scholar can continue
studying for many years. The practice of young boys attending a priest school
for a couple of years before entering the state system or dropping out may have
diminished but is still common in rural areas.

Previous State Education Language Policy

The first state school was established in 1908, primarily for the promotion of
foreign languages, with obligatory French, and optional English, Italian and
Arabic. This was an instrumental measure for ‘the speedy provision of a cadre
of young men who could cope with foreign affairs’ (Ministry of Education and
Fine Arts, 1965: 15). (The school is still extant as a conventional secondary school.)
Amharic was not taught as a subject until 1919, when one teacher of Giiz and
Amharic was assigned to each of the two existing schools. Haile Selassie
repeatedly promoted Amharic in public pronouncements as ‘the learning of our
own country’ and, addressing students, ‘the reading and writing of your own
country’s language (Amharic)’ (Ministry of Education and Fine Arts, 1964:
16–17). From 1947 to 1958, the medium of instruction in all schools was English,
with Amharic taught as a subject, but in 1958 the primary-school syllabus
switched to Amharic as the medium with English as a subject, English remaining
the medium in secondary school. The implementation was subsequently
 Issues in Ethiopian Languag e Policy and Educ ation 327
condemned as being over-hasty and ill-prepared, particularly with regard to the
production of textbooks and related materials (Tesfaye, 1976).
The elementary school curriculum of 1963 explicitly stated that it was ‘six years
of study to be taught completely in Amharic with English {as FL} taught as early
in the programme as possible’ (Tesfaye, 1976: 391). This was spelt out as
involving some delay with English for non-native speakers of Amharic, extra
attention being paid to Amharic. In practice though, little, if any, systematic
provision was ever made for students with other mother tongues.
In secondary schools since 1941, the medium of education has always been
English, except in Amharic lessons and morals (that is, religious education,
replaced by political education under the Dergue). French was an optional
subject, available in the Ethiopian School Leaving Certificate along with the
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

non-optional English and Amharic. No indigenous language other than Amharic

was taught in the school programme at any level. The inclusion of Oromo as well
as Amharic terms for trees and crops in the excellent English-medium official
school geography book (Last, 1954) was exceptional.
Under the Dergue (1974–91) there was little effective change in language
policy although some attention was paid to vernacular languages. Amharic
remained the official national language, in effect a symbol of national unity and
the medium of primary education, with English as a semi-official second
language of limited scope and the medium of secondary and higher education.
Table 2 indicates the prevailing pattern prior to 1991. It is important not to lose
sight of the fact that only a minority of Ethiopian citizens receive any formal
education, and of those who do, 40% do not continue to secondary level, many
dropping out at Grade 1 or 2 and relapsing to illiteracy. There appear to be
significant improvements in primary-school recruitment, from 19.2% in 1984 to
35% in 1993 (National Office of Population, 1993) but these statistics say nothing
about how many children complete primary education. According to recent
figures, 20% of the appropriate age group attend primary and only 12%
secondary school (Ministry of Education, 1993). Nevertheless, although it is
highly competitive, state education has surprisingly never been exclusively a
monopoly of the rich because it has always been free; and figures will improve
if political stability is maintained, though there is some talk of requiring students
to contribute to the cost of education. Seventy-five per cent of the 20% attending
Table 2 Old curriculum
1
Level Years Medium
2
University 4 English
3
Secondary 4 :Grade 9-12 English
3
Junior secondary 2 :Grade 7-8 English
Primary 6 :Grade 1-6 Amharic (English as subject)

1. Where the language is the medium, it is also timetabled as a taught subject (except in later
stages of university).
2. All faculties except engineering (5 years) and medicine (7).
3. De facto, Amharic sometimes unofficially used as a medium because of comprehension
difficulties in English.
 328 Journal of M ultiling ual and M ulticultural D evelop m ent
primary school do so in urban areas. Some of these are rural children who have
moved into the towns in order to be educated, but they rarely return to the
countryside. Secondary education is overwhelmingly in medium or large towns.
Of the 11,000 secondary teachers, only 40% were deemed by the Ministry of
Education (1993) to be suitably qualified.

The New Language Policy

In 1991, the EPRDF toppled the Dergue and took power, and Eritrea became
an independent state. Throughout the revolutionary struggle, language had
always been an issue, reflecting the somewhat ethnic orientation of the main
opposition groups (though this is an oversimplification). The Eritrean anti-
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

Dergue forces had pursued a literacy programme in Tigrinya, as had the Tigray
People’s Liberation Front, and the Oromo Liberation Front had done some
literacy work in Oromo. The traditional Ethiopian government tendency to play
down linguistic differences and ethnic groupings gave way to much overt talk
of bihersibocc (nationalities).
A policy decision issued on 20th July 1991, and the draft Constitution of 1993,
divided the country into nine regions based on the majority languages spoken in
the areas concerned — henceforth nationality languages or vernaculars. The only
major surprise is the inclusion of Arabic for the Benishangul region bordering
Sudan. The new policy became official in 1994. Not surprisingly, it placed
considerable emphasis on languages in education, with changes affecting the
roles of Amharic and English and, for the first time, incorporating other
indigenous languages in the school curriculum. Amharic retained its official
status as national language but not its role as medium of education at primary
level (except, of course, in regions where Amharic is the designated ‘nationality
language’). The new policy decreed that::
(1) Primary education will be in the medium of the nationality language of the
region. This will also be the medium of teacher training institutes for
kindergarten and primary education.
(2) Amharic will be taught as a language of country-wide communication.
(3) English will be the medium for secondary and higher education.
(Transitional Government of Ethiopia, 1994: 23.)
Table 3 indicates the new programme. The change in the structure of the school
programme pushes what used to be the two junior secondary years (Grades 7
and 8) into the primary school, replacing two years of English-medium with
vernacular-medium. Otherwise English retains its status as the medium for
secondary and higher education. Amharic, however, is drastically affected
(except in Amharic vernacular regions), since it is reduced from being the sole
medium throughout primary school, as well as a timetabled subject throughout
primary and secondary, to merely being a subject in upper primary and
secondary (Grades 5–12).
Appendix 1 (using categories devised for Tanzania by Rubagumya, 1986)
indicates the effects of the new language policy envisaged for a range of social
domains. The most striking feature is the upgrading of vernaculars at the expense
of Amharic. Of all the recognised ‘nationalities’, only the Afar (largely nomadic
 Issues in Ethiopian Languag e Policy and Educ ation 329

Table 3 New curriculum

1
Level Years Medium
University 42 English
Secondary 2: 2: Grade 11-12 English
Senior high school
Secondary 1: 2: Grade 9-10 English (Amharic as subject)
General education
Primary 2: 4: Grade 5-8 Vernacular (English as subject)
General education
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

Primary 1: 4: Grade 1-4 Vernacular (English as subject)

Basic education

1 Where the language is the medium, it is also timetabled as a taught subject (except in later
stages of university).
2 All faculties except engineering (5 years) and medicine (7).

pastoralists in the deserts of the north-east) have opted for Amharic as medium
in preference to their own Cushitic tongue.
A very controversial aspect of the new policy is the graphisation of the
vernaculars. The trend seems to be that all except Amharic and Tigrinya will
adopt the Roman alphabet, but at the time of writing there is ongoing debate
about this. The Ethiopic writing system, which can be found in inscriptions from
the fourth century, was bequeathed by Giiz to Amharic and Tigrinya and, until
recently, was the obvious candidate for establishing literacy in the vernaculars.
In origin, it is a South Semitic script, probably derived from the Sabean system.
It essentially uses one character per syllable, and consists of 275 symbols. To
simplify somewhat, the system works on the basis of seven ‘orders’ for each basic
consonant symbol, representing seven vowels; in addition there are some
additional variants for labialised consonants (with vowel). For historical reasons,
there is some redundancy of symbols. Moreover, there are some ambiguities,
though far fewer than in English orthography. Spelling conventions are
standardised to some degree, but there is sometimes uncertainty over logically
possible alternatives; even so, spelling is far more predictable than in English
and, for the native speaker, reading is fairly straightforward.
At present, virtually all literate Ethiopians are literate in this system; far fewer
are literate in the Roman alphabet and very few in Arabic script. Any Ethiopian
who is literate in the Roman alphabet is normally literate in the Ethiopic system
though the reverse, of course, is not true. This is the first point in the history of
Ethiopia where the predominance of this script has been challenged.

Discussion
There seem to be clear social benefits in the new policy, which is in line with
current progressive thinking in its respect for ethnic sensibilities and the use of
the mother tongue in initial education. United Nations policy, for example, has
 330 Journal of M ultiling ual and M ulticultural D evelop m ent
long subscribed to the latter principle (though not always consistently): ‘every
child has the right of instruction in his {sic} own mother tongue’ (UNESCO, 1953).
The ‘nationalities’ issue has been one of the major factors in the civil wars of
the past 30 years. Considered in terms of sentimental as opposed to instrumental
interests, at least, the new policy can be viewed as a marked improvement on the
Amharic linguistic hegemony of the past. In political terms, moves in this
direction are virtually inevitable, given the EPRDF’s claims of commitment to
decentralisation and egalitarian treatment of nationalities. As Nadkarni says:
The problem of language planning for multilingual societies is basically one
of bringing about national cohesiveness and of creating a national identity
without, at the same time, displacing the original ethnic-cultural identities
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

of the component groups. (Nadkarni, 1984: 152)

This is probably the first Ethiopian government that has given much thought
to the last proviso, and it is ostensibly an attractive development. However, there
are significant drawbacks at the practical level.
First of all, there is no possibility of implementing the UNESCO principle of
mother-tongue teaching for ‘every child’. Even assuming universal education,
which is a long way off for Ethiopia, this ideal is virtually inconceivable in a
country with such a degree of pluralism (cf. Bloor & Bloor, 1990; Goyvaerts et al.,
1983). If there are 99 (or 70, or whatever) languages and only nine linguistic
regions, it is evident that many minority languages will not be accommodated
(though some regions may recognise more than one language). Many speakers,
even of one of the major languages, may be obliged to receive their primary
education in some other language because many areas, especially towns, are
linguistically heterogeneous. This inconvenience is not in itself an argument
against increased recognition of pluralism, since a partial treatment is often better
than no treatment at all, but it is a point to bear in mind.
Sociolinguists tend to prefer gradual to sudden change; politicians, especially
in revolutionary situations, prefer the grand sweep in which ‘a decision is taken
at the moment of planning that a complete reconstruction of some kind is possible
and desirable’ (Corson, 1990: 142–3). Corson notes that this type of planning is
fraught with risks, and suggests a slower approach. Similarly, Kennedy argues:
Language change should be phased, move at a speed commensurate with
social acceptance and be made in line with social trends, not by decree,
otherwise community antagonism will prevent implementation. (Kennedy,
1982: 273)
It is hard to estimate what the degree of social acceptance of the new policy is.
There is probably considerable support in many quarters — though first-
language Amharic speakers are unlikely to perceive many advantages, since they
have always had the convenience of initial mother-tongue teaching, regardless
of where they lived: an advantage they will now lose. But, within the school
system, it may well be non-Amharic speakers who miss out by their relative lack
of exposure to what remains the official language for many internal purposes,
the main lingua franca and the most well-established indigenous language. It is
possible that under the new dispensation some of these strengths of Amharic will
 Issues in Ethiopian Languag e Policy and Educ ation 331
diminish, but it is not clear how this will benefit the nation as a whole, especially
as English will now relatively dominate the education system more than ever.
Attitudinal factors have to be taken into account, and it is indisputable that
individuals perceive English as highly desirable, even those who do not speak it (see
Cooper & Singh, 1976, on Ethiopian factory workers). Even Phillipson (1992)
concedes that individual attitudes in Africa and elsewhere may favour English,
seeing it as a means of personal advancement, and this is certainly the case in
Ethiopia. He makes a forceful, extended attack on the ‘linguistic imperialism’ of
English, but the situations that he discusses are primarily ex-British (or American)
colonies. Although there are obvious neo-colonialist factors, English in Ethiopia is
free of much of the colonial baggage that it carries in Anglophone Africa. However,
the Ethiopian situation does resemble that of, say, Kenya or Nigeria (see Mann, 1991)
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

insofar as English has a neutrality with regard to ethnic groups in Ethiopia that
Amharic, the only serious contender, lacks. As Schmied observes:
Ethnic languages are normally not accepted as national languages wherever
other groups fear ‘tribal dominance’ and prefer English, which is ‘tribally
neutral’. Only tribally neutral lingue franche have any chance of taking over
certain functions from English as national languages. (Schmied, 1991: 27)
Once again, though, Schmied’s paradigm is of ex-British colonies, and English
is a weak option for Ethiopia. The difference here is that it is not a question of
‘taking over’ from English since Amharic was already in situ and English never
has been. Outside the education system, the limitations of English in the
Ethiopian situation are obvious. It is barely conceivable that it could ever fulfil
the functions that Amharic has fulfilled in the past (and this is unlikely to be the
planners’ intention). It is a foreign, rather than a second, language with nothing
like the scope and historical importance it has in most ex-colonies. A monoglot
English-speaker can cope in a limited way in Addis Ababa and one or two other
urban centres, but elsewhere English has very limited potential. It is true that
there are regions where Amharic is not the preferred lingua franca, for example
in Tigray where Tigrinya dominates, or in Wollega province, where Oromo is the
optimal second language as well as the majority first language. Nevertheless,
Amharic has always had a role even there.
Furthermore, there has been considerable dissatisfaction over the years with
the standards of English attained by students, which led to the original
replacement of English as a medium at primary level by Amharic, and to serious
consideration of abandoning English as a medium altogether (Tesfaye, 1976;
Rogers, 1982). It is probably easier to attain widespread high standards of
Amharic than of English. Perhaps it is inevitable that English should remain the
medium at higher levels, but the disadvantages of down-grading Amharic in
favour of English are substantial.
Indeed, there is a very strong instrumental case for expanding rather than
reducing the importance of Amharic. On Nida’s (1971) tripartite classification of
language uses, the obvious allocation of communication roles is:
English: international
Amharic: national
Vernacular: local
 332 Journal of M ultiling ual and M ulticultural D evelop m ent
This is what the new policy explicitly says (see above), but it seems inconsistent
that Amharic has such a small place in the school curriculum compared to English
and the vernacular (where this is not Amharic). Phillipson and Skutnabb-Kangas
include in their analysis of the meaning of linguistic human rights not only ‘the
right to learn the mother tongue’ but also ‘the right to learn at least one of the
official languages in one’s country of residence’ (1995: 488). It would be a pity if,
for most Ethiopians fortunate enough to receive schooling at all, this turned out
to be English. Moreover, the high drop-out rate (or ‘push-out’, as Phillipson, 1992,
puts it) means that even fewer people will have access to this important lingua
franca, a fact which must have implications for national unity as well as social
mobility. Non-Amharic speakers who do not go beyond the first stage of primary
(Grades 1–4) will be cut off from Amharic, and thereby from potential access to
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

a large and sophisticated indigenous written literature. Of course, there is a

wealth of oral literature in the vernaculars and no doubt, as graphisation
develops, this will be written down and new written work will emerge. All of
this is highly desirable, but it will be a long time materialising. Meanwhile, an
already available and impressive cultural heritage may be diminished.
It is understandable that non-Amharas should react against the advantages
long enjoyed by speakers of Amharic, and there are obvious instrumental as well
as sentimental benefits in upgrading other vernaculars, but there is a massive
cost, especially as the downgrading of Amharic probably entails the upgrading
of English. One incidental danger is that of a rift based on knowledge of English
between the highly educated and the rest (consider Czarist Russia and French).
This danger was already present in the old order, but it could be intensified by
the increased emphasis on English at the expense of Amharic. Phillipson (1992)
draws attention to the Organisation of African Unity’s stipulation in its charter
(1963) that the official use of foreign languages would be only provisionally
tolerated, and that the use of African languages would be encouraged.
Phillipson’s point is that this promise has not been realised, but he cites as
exceptions to European language dominance Somali in Somalia and Swahili in
Tanzania. Amharic would have been an equally good example, but it may not
3
remain so.
It is interesting to note that UNESCO itself dropped the principle of vernacular
initial education in Rhodesia in 1964 in the face of what they saw as the confusing
sequence: mother tongue vernacular — official vernacular — English; concluding
that English should be the medium throughout (Schmied, 1991: 106). This
conclusion is certainly open to serious challenge, even though the potential of
English in what was then Rhodesia was (and still is in Zimbabwe) markedly
greater than it is in Ethiopia, but it indicates the gap between ideological
principles and their implementation.
There are also serious practical difficulties in the implementation of the switch
to vernacular medium, difficulties familiar from other discussions of language
planning in pluralistic societies (for example, Todd, 1984; Mann, 1991). There is
a total lack of educational materials in most Ethiopian mother tongues, and, more
than that, there is a lack of any kind of written material whatsoever since most
languages have scarcely, if ever, been written. (Where they have been, the script
used has normally been Ethiopic.) This deficit is being met by the hasty
 Issues in Ethiopian Languag e Policy and Educ ation 333
translation of basic texts into the vernaculars by teachers whose sole qualification
for the job is that they can speak the vernacular. Few lessons, it seems, have been
learned from the unsatisfactory outcome of similar hasty procedures for the
arguably less difficult Amharicisation of the school programme in 1958.
There may well also be insufficient numbers of teachers who can teach in the
vernaculars — or in some cases even speak them fluently. And, as Todd says:
If automatic retrieval of knowledge in a language other than the one in
which it was acquired is not possible in a highly trained individual who is
bilingual in two similar languages, at home in both cultures, then we must
be wary of assuming that it is easy or even possible to learn in a world
language and teach the same material in a non-related vernacular. (Todd,
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

1984: 166)
The new policy in Ethiopia sensibly takes account of such reasoning by making
teacher education take place in the vernacular. However, it will take several years
to train these teachers (and who will train the trainers?) and the change of
medium in schools is already being implemented. Moreover, the problem of
teacher mobility will be seriously aggravated since the geographical scope will
be diminished, and, given the already grave deficiency in teaching provision, this
is not a trivial consideration. This is not to say that vernacular primary teaching
should be rejected, but it does raise questions about the pace of change and the
manner of its implementation. To see the scale of the problem, compare the
difficulties experienced even in affluent societies with a comparatively simple
language problem, such as Norway with its implementation of Sámi education.
(Corson, 1995.)
The choice of writing systems for the vernaculars raises considerable
difficulty, yet, as mentioned above, all except Amharic, Tigrinya and Afar seem
to have decided on Roman script rather than Ethiopic. As with many of these
developments, the choice is no doubt strongly motivated by ethnic political
considerations. The revolution has been partially characterised as struggle
against an Amharic-dominated system with the intention of securing equality
for other ‘nationalities’. It may partly be the desire to distance themselves from
Amharic that inspires this near unanimity; or, more subtly, it may be due to a
concern to be seen to establish some distance.
There are obvious practical attractions in the Roman alphabet. The memory
load required for around 26 symbols plus a few diacritics is ostensibly
considerably less than that required for the Ethiopic system with more than ten
times as many characters. Since the orthographies are starting free from existing
conventions, it should be possible to establish a reasonably phonetic system, as
we see from the example of Turkish. Access to English and many other external
languages could be facilitated, and internationally available equipment for
typing, printing, and word processing can be used with little modification.
The arguments against this choice, however, are at least as strong. The contrast
between 26 letters and over 270 syllabic signs is misleading: the regularities of
Ethiopic script significantly diminish the memory load. The Ethiopic script is well
established throughout the country, through Giiz in the Church and Amharic as
the national language. Literacy programmes in the past have always taught the
 334 Journal of M ultiling ual and M ulticultural D evelop m ent
Ethiopic system, and those early school-leavers who have retained literacy to
date have done so in that system. Official reports have been made of a literacy
rate of 62.4% compared with 7% in the early 1970s (cited in National Office of
Population, 1993). This startling claim begs a few questions, but we can accept
that there has been a great increase in literacy and that it has been through the
medium of Ethiopic rather than Roman script. Ethiopic is at present incompara-
bly better known than the Roman alphabet, and typewriters, printing presses,
and so on, already exist, including word processing software for converting from
Roman.
It is particularly surprising that Ethio-Semitic languages such as Gurage
should adopt the Roman alphabet, since the acquisition of written Amharic for
Gurages (many of whom can speak it already) would presumably be relatively
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

easy if the same script were used. There is already a small amount of literature
published in Gurage, using the Ethiopic script. It is in some ways comparable
with the unimaginable scenario of Catalan adopting a different script from that
4
used for Spanish. Oromo has also made considerable use of this script, including
the publication of a translation of the Bible. Somali is a more difficult case as it is
not confined to Ethiopia; there is also the (numerically small) problem of Adare,
spoken mainly in and around one of Ethiopia’s largest cities, Harar, which has
already used Arabic script to some extent and, more strikingly, there is the new
recognition of Arabic for Beneshangul.
Public written phenomena known to require and reinforce functional literacy,
such as place-name signs, advertisements and maps will need to be written in
two scripts even when they are language-neutral, adding to material cost.
Personal names, signatures, and so on, will confusingly involve different scripts.
More importantly, students in non-Amharic regions who leave school before the
second phase of primary (i.e. after four years or less) will have no exposure to
the Ethiopic system and, so far at least, there have been many such early leavers.
Work-force mobility for employment requiring limited functional literacy will
thus be seriously curtailed.
In principle, one possible alternative is for all the languages, including
Amharic and Tigrinya, to adopt the Roman system, but this is hardly more likely
than the Arab world doing so. The Ethiopic script is central to the culture of
Amharic and Tigrinya, and has long played a part in the culture of other ethnic
groups, not least in their contribution to discussions of reform. The special status
of Giiz as a religious and literary resource is intimately bound up with the script
in which it is written. Although it is the practical considerations that are most
persuasive, it is also true that the script’s emotive religious and aesthetic
5
connotations rule out its disappearance. Even if there were no instrumental basis
for its retention, which is certainly not the case, the sentimental considerations
would be overwhelming. The change from Arabic script to Roman in Atatürk’s
Turkey offers a precedent for such a dramatic change, but Arabic script was not
unique and distinctive to Turkey, as Ethiopic is to Ethiopia.
Thus, if the Roman system is adopted for Oromo and other languages,
Ethiopia will be committed to two writing systems for indigenous languages (or
three, including Arabic). Whilst Amharic remains the most accessible second
language, this considerably reduces the possibility of widespread literacy in
 Issues in Ethiopian Languag e Policy and Educ ation 335
more than one indigenous language, even for bilinguals and multilinguals, of
whom there are many.

Conclusion
The spirit of the age seems to favour a move away from large highly
centralised nation states towards regionalism, federalism or balkanisation, and
Ethiopia is no exception to this trend. True, in Ethiopia as in other pluralistic
societies, ‘a policy guaranteeing and implementing a measure of equal treatment
for languages {¼} is a natural step to take’ (Corson, 1990: 142). It is also no doubt
equitable and in many ways desirable but, in terms of its full realisation for all
languages in the country, it is not feasible for a long time to come, if ever; and for
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

its partial realisation, the difficulties are enormous and the financial and social
costs considerable. Ethiopian unity might well be a casualty in the long term. On
the other hand, Phillipson & Skutnabb-Kangas are almost certainly right when
they say: ‘Linguistic diversity is not causally related to conflict though of course
language is a major mobilising factor in contexts where an ethnic group feels itself
threatened’ (1995: 482). A happily tolerant pluralistic society might be achieved
by the recognition of the differences within it, and the threat of going the way of
Yugoslavia and many other such societies thereby averted. This will take place
only in the context of feasible and carefully implemented language planning for
education and for society at large.

Notes
1. In line with Tigrinya, we might have used the indigenous term Amharinya, but we have
chosen to follow the standard practice in English of calling this language Amharic. We
have used the terms Adare, Afar, Gurage, Oromo, Somali for languages and for peoples.
We also follow the current Ethiopian practice of calling the Oromo by their own name
rather than by the now obsolete term Galla; similarly, we speak of the Afar rather than
the Danakil or Denkali.
2. All these figures are rounded to the nearest decimal point.
3. The comparison of Amharic to Somali or Swahili is tenuous, of course. Somalia is
relatively culturally and linguistically homogeneous, and Swahili has little, if any, of
Amharic’s connotations of being the language of a traditionally dominant group.
4. One big difference, though, is the existence of a long-written literary tradition in
Catalan.
5. This factor might conceivably be a negative one for the Muslim population.

References
Bender, M.L., Bowen, J.D., Cooper, R.L. and Ferguson, C.A. (eds) (1976) Language in
Ethiopia. London: Oxford University Press.
Bloor, M. and Bloor, T. (1990) The role of English in resurgent Africa. In R. Clark and N.
Fairclough (eds) Language and Power (pp. 32–43). London: Centre for Information on
Language Teaching and Research for the British Association for Applied Linguistics.
Cooper, R.L. (1976) Government language policy. In M.L. Bender, J.D. Bowen, R.L. Cooper
and C.A. Ferguson (eds) Language in Ethiopia (pp. 187–90). London: Oxford University
Press.
Cooper, R.L. and Singh, B.N. (1976) Language and factory workers. In M.L. Bender, J.D.
Bowen, R.L. Cooper and C.A. Ferguson (eds) Language in Ethiopia (pp. 32–43). London:
Oxford University Press.
Corson, D. (1990) Language Planning across the Curriculum. Clevedon: Multilingual
Matters.
 336 Journal of M ultiling ual and M ulticultural D evelop m ent
— (1995) Norway’s ‘Sámi Language Act’: Emancipatory implications for the world’s
aboriginal peoples. Language in Society 24 (4), 493–514.
Edwards, J. (1994) Multilingualism. London: Routledge.
Ferguson, C.A. (1967) National sociolinguistic profile formulas. In W. Bright (ed.)
Sociolinguistics (pp. 309–24). The Hague: Mouton. (Also in A.S. Dil (1971) Language
Structure and Language Use: Essays by Charles A. Ferguson (pp. 157–85). Stanford:
Stanford University Press.)
— (1970) The role of Arabic in Ethiopia: A sociolinguistic perspective. In Georgetown
University Monograph Series on Languages and Linguistics: Monograph No. 23 (pp. 355–68).
(Also in A.S. Dil (ed.) (1971) Language Structure and Language Use: Essays by Charles A.
Ferguson (pp. 355–68). Stanford: Stanford University Press).
Goyvaerts, D.L., Semikenke, M.W.K and Naeyaert, D. (1983) Language and education
policy in the multilingual city of Bukavu. Journal of Multilingual and Multicultural
Development 4 (1), 47–62.
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

Habte-Mariam Marcos (1976) Three other Ethiopian languages: (3) Italian. In M.L. Bender,
J.D. Bowen, R.L. Cooper and C.A. Ferguson (eds) Language in Ethiopia (pp. 170–82).
London: Oxford University Press.
Kennedy, C. (1982) Language planning. Language Teaching 15 (3), 264–84.
Khubchandani, L.M. (1984) Language planning processes for pluralistic societies. In C.
Kennedy (ed.) Language Planning and Language Education (pp. 98–110). London: George
Allen and Unwin.
Last, G.C. (1954) Our Wealth. Addis Ababa: Ministry of Education with Ethiopia-United
States Communications Media Co-operative Service.
Mann, C. (1991) Choosing an indigenous official language for Nigeria. In P. Meara and
A. Ryan (eds) Language and Nation (pp. 91–106). London: British Association for
Applied Linguistics in association with Centre for Information on Language Teaching
and Research.
Ministry of Education (1993) Ethiopian Educational Policy Implementation Strategy. Addis
Ababa: Ministry of Education.
Ministry of Education and Fine Arts (1964) Project for Expansion of Second Level Education
in Ethiopia. Addis Ababa: Ministry of Education.
— (1965) Curriculum Policies and Standards: Teacher Training Institutes. Addis Ababa:
Ministry of Education.
Nadkarni, M.V. (1984) Cultural pluralism as a national resource: Strategies for language
education. In C. Kennedy (ed.) Language Planning and Language Education (pp. 151–59).
London: George Allen and Unwin.
National Office of the Population and Housing Census Commission (1987) Population and
Housing Census of Ethiopia, 1984: Analytical Report on Results for Addis Ababa Vol. 1, No. 1.
National Office of Population (1993) National Report on Population and Development. Addis
Ababa: National Office of Population, Office of the Prime Minister.
Nida, E. (1971) Communication roles of language in multilingual societies. In A. Dil (ed.)
Language Structure and Translation (pp. 147–73). Stanford: Stanford University Press.
Phillipson, R. (1992) Linguistic Imperialism. Oxford: Oxford University Press.
Phillipson, R. and Skutnabb-Kangas, T. (1995) Linguistic rights and wrongs. Applied
Linguistics 16 (4), 483–504.
Rogers, J. (1982) The world for sick proper. English Language Teaching Journal 36 (3), 144–51.
Rubagumya, C.M. (1986) Language planning in the Tanzanian educational system. Journal
of Multilingual and Multicultural Development 7 (4), 47–62.
Schmied, J. (1991) English in Africa: An Introduction. Harlow: Longman.
Senayit Lemma (1994) Yekwankwa attekakemina sirichit bAddis Ababa (Language
distribution and use in Addis Ababa). Unpublished paper presented to the Second
Language Symposium at Kotebe College of Teacher Education.
Tesfaye Shewaye (1976) Language Curricula: (1) Amharic. In M.L. Bender, J.D. Bowen,
R.L. Cooper and C.A. Ferguson (eds) Language in Ethiopia (pp. 371–80). London: Oxford
University Press.
Todd, L. (1984) Language options for education in a multilingual society: Cameroon. In
 Issues in Ethiopian Languag e Policy and Educ ation 337
C. Kennedy (ed.) Language Planning and Language Education (pp. 160–71). London:
George Allen and Unwin.
Transitional Government of Ethiopia (1994) Education and Training Policy. Addis Ababa:
Transitional Government of Ethiopia.
UNESCO (1953) The Use of Vernacular Languages in Education. Paris: UNESCO.
Wedekind, K. (1994) Updating linguistic maps. In Survey of Little Known Languages
Linguistic Reports No. 13. Addis Ababa: Institute of Ethiopian Studies/Summer
Institute of Linguistics project.
Downloaded by [University of Auckland Library] at 16:46 09 April 2015
 338 Journal of M ultiling ual and M ulticultural D evelop m ent

Appendix Approximate sociolinguistic domains for Ethiopia before and after

Draft Constitution 1993
Previous New plan
Domain V A E O V A E O
1 INFORMAL
in the home + (+) +
with neighbours + (+) +
in work place + (+) + (+)
2 CULTURAL
at place of worship (+) + +1 + +1
written literature {+} + {+} + (+)
Downloaded by [University of Auckland Library] at 16:46 09 April 2015

cinema {+} + {+} ? ? +

3 COMMERCIAL
big business (+) + {+} + +
small business + + 2 + (+)
tourism + {+} ? +
4 EDUCATIONAL
primary ed. medium + +
2ry & 3ry ed. medium + +
adult ed. medium + +
books, journals etc. + (+) +
5 POLITICAL
parliament + +
public rallies (+) + + (+)
6 ADMINISTRATION
village + (+) +
district/regional + +
national + +
7 JUDICIARY
primary court + +
district court + +
high/appeal court + + (+)
8 MASS MEDIA
radio {+} + {+} (+) + (+)
television (very limited) + + (+) + (+)
daily papers + {+} (+) + ?
9 INTERNATIONAL
diplomacy + (+) + (+)
trade + (+) + (+)
cultural exchange + (+) + (+)
information exchange + +
science and technology + +

Domains adapted from Rubagumya (1986)

Key: V: Vernacular; A: Amharic; E: English; O: Other; + normal in most contents; (+) in limited
contexts; {+} in rare contexts; 1: restricted Giiz for Orthodox Christian liturgy; restricted classical
Arabic for Mosque (Note Arabic is a vernacular under the new plan)