Daniel S. Alemu Planning and Changing Abebayehu A. Tekleselassie Vol. 37, No. 3&4, 2006, pp. 151–168 INSTRUCTIONALLANGUAGE POLICYIN ETHIOPIA: MOTIVATED BYPOLITICS OR THE EDUCATIONAL NEEDS OFCHILDREN? Introduction Following the change of government in 1991, Ethiopia’s educa- tion system has been undergoing fundamental change. No part of the edu- cation system has been left untouched. Acentralized administration was replaced by a decentralized one in line with the principle of federalism under which the current government has been organized. The federal Min- istry of Education, which was highly centralized in the past, has now been decentralized into many states, district-level bureaus, and departments demarcated on ethnic lines. The decentralization process, in addition to devolving authority, has brought with it various change initiatives, one of which is the change in media of instruction. Until the current government took power, the media of instruction in Ethiopia’s formal education system were Amharic (for elementary level) and English (for junior high and above). Whereas the socialist gov- ernment (1974–91) had encouraged the use of some 15 ethnic languages in non-formal education, the imperial regime (which ruled the country until 1974) preferred to use one official language (Amharic) with the intention of safeguarding national integrity. According to the 1994 census, more than 80 ethnic groups exist in Ethiopia. The new ethnic-based states were demarcated into 14 (at least initially) ethnic-based boundaries that comprise as many as 20 ethnic groups per state. The number of languages used as media of instruction varies from state to state. In the Southern Nation, Nationalities and Peo- ples Regional State (SNNPRS) alone, for instance, eight local languages of instruction have been in use at the primary level (Cohen, 2000). The authors of this work appreciate the advantages of vernacular education for children but argue that rushing to formulate and to imple- ment the new instructional language policy, without considering such fac- tors as the existing unbalanced level of development among the newly created states and the absence of any meaningful preparation to deliver education in the newly chosen languages, appears to be more of a political gimmick than a sound pedagogical move. The current study intends to explain the formulation, implementation, and outcome of Ethiopia’s instructional language policy in light of the PRINCE system of power analysis as adapted by Fowler (2004) along with several literature refer- ences pertinent to the issue. After providing a brief background on Ethiopia and its education and language of instruction policies, this article analyzes the formulation and implementation of Ethiopia’s present instructional language policy and problems therein. 151 Alemu Tekleselassie Background Ethiopia According to the census data of 1994 (the most recent year for which data are available), Ethiopia has a population of 63 million, the sec- ond largest in Sub-Saharan Africa. About 83% of the population lives in rural areas engaged in subsistence agriculture. As per the census, the adjusted total fertility rates are 6.74 for the country, 4.5 for urban areas and 7.2 for rural areas. The estimated infant mortality rate for the country is 11.6%. The primary school age population (7–14) constitutes 12.6 mil- lion, while the secondary school age population (15–18) totals 5.58 mil- lion. About half of the population is female. Religion is one of the socio-cultural characteristics of the popula- tion. According to the census, the composition by religion shows that 50.6% are Orthodox Christian, 32.8% Muslim, 10.2% Protestant, 4.6% followers of traditional religion, and 1.6% constituted by other religions. Ethnic diversity is one of the most important attributes of the pop- ulation. The 1994 census data identified over 80 ethnic groups in the country. The distribution of the major ethnic groups and their proportions in the population are provided in the Appendix. Education in Ethiopia Ethiopia is unique from other African countries in that it was never colonized by any foreign power except a five year invasion by Italy from 1936 to 1941. An ancient nation, with the legend of the Queen of Sheba, Ethiopia “may have had its origin in the early period of Sabean migration to Africa” (Wagaw, 1979, p. 2). Also, Ethiopia is one of the few countries in the world with a long-standing literary history (Tekeste, 1996) and with its own scripts still actively in use. The country’s current educational status, however, is strikingly disappointing even at the African standard. According to the 2002/03 educational annual abstract of the Ministry of Education, the gross enrollment ratio at primary and secondary school levels respectively was 64.4 and 19.3 percent. The enrollment ratio in higher education stands below 1% of the expected age group (Ministry of Education, 1999). Low as the enrollment rate has been, severe gender and regional disparities further compound it. While women make up 50% of the popu- lation, they represent only 41.2% of primary, 36.6% of secondary, and 20.5% of tertiary level enrollment (Ministry of Education, 2003). The regional disparity in primary school gross enrollment rates (which include over-age students) ranges from the lowest of 13.8% to the highest of 135.4% (Ministry of Education, 2003). Variations are also large between rural and urban areas. Among the several reasons for Ethiopia’s low educational devel- opment and its impact on the socioeconomic development of the country, the major ones are harsh treatment with socialism for seventeen years and periodic war and famine (Cummings, 1999). Nationals by and large also 152 Planning and Changing Instructional Language Policy in Ethiopia blame the country’s dismal educational record on unsound policies of suc- cessive governments. It is partly and thusly against this backdrop that the new policy was envisaged. The Policy on Language of Instruction in Ethiopia: Values and Philosophies Despite the fact that Ethiopia is a multilingual country, a single language (Amharic) had been the medium of instruction at the elementary school level throughout much of the country’s history. It was only in 1974, when the socialist government came to power, that the use of ethnic lan- guages (also called “nationality languages” since the current government uses the latter terminology synonymously with “ethnic languages” in its official documents) for instructional purposes was considered as an issue. Why did it take such a long time for ethnic languages to become a policy issue in Ethiopia? One may find part of the answer for this question in the country’s history, and mainly in its system of government. For several cen- turies, Ethiopia had been under a feudal monarchy. It was thus quite incon- ceivable for the imperial regime to address ethnic issues that are enshrined in democratic values. The government’s determination to bury ethnic lan- guages out of the policy agenda may also correspond with the country’s long history of independence. Successive imperial regimes advocated for the use of one national language as an instrument for maintaining the coun- try’s integrity. Introducing other languages for instructional purposes had been conceived as courting national disintegration. In all cases, the imperi- al regime was not longsighted enough to perceive the danger of imposing one national language on the multiethnic nation, which was like “a defacto declaration of war on the others” (Seyoum, 1997, p. 2). The socialist government that assumed the mantle of leadership in 1974 shifted from promoting one language as an instrument for national unity to encouraging the use of other languages as per its political orienta- tions. One notable effort during this time was the policy decision to con- duct adult literacy programs in fifteen ethnic languages (Ayalew, 1999). The literacy program (campaign) started in 1979 and ended only at the fall of the socialist government in 1991. The other noteworthy policy decision by the socialist regime was on transcribing these languages in the Ethiopic script (traditionally used for Semitic languages in the country) most of which were in unwritten form hitherto. However, the use of these lan- guages was limited to the non-formal education sector and the govern- ment did not push forward to use them as instructional languages in the formal system (Ayalew, 1999). Hence, Amharic (a Semitic language) remained as the only national language that was used as the medium of instruction for formal education at the elementary level. When the current government came to power in 1991, a potential condition was created for ethnic languages to reemerge as a major policy issue. Two factors, among others, accentuated the need for this change. The first was the political orientation of the government. Having replaced the totalitarian socialist regime, the government shifted to introduce a kind of “Western Democracy” and its accompanying values. Liberty, equality, jus- Vol. 37, No. 3&4, 2006, pp. 151–168 153 Alemu Tekleselassie tice, truth, and respect for human rights became the agenda of the govern- ment (Seyoum, 1996). The second factor was the unparalleled premium placed by the ruling party, the Ethiopian People’s Revolutionary Democra- tic Front (EPRDF), on ethnic-based politics in the country’s history. Main- ly representing the Tigraway (Tigrian) ethnic group, EPRDF labeled the Amhara ethnic group as the suppressor and all the non-Amhara ethnic groups as the suppressed whose languages, traditions, and cultures had been subjugated (Seyoum, 1996). In its attempt to redress such inequali- ties, the government vowed the issue of ethnic languages to be its top agen- da and policy priority. To that effect, one may also argue that it was an opportune time for the issue to be addressed. What requires inquiry then is how this laudable issue has been approached (i.e. the policy process). The New Instructional Language Policy Policy Formulation At various points in the history of Ethiopia, political motives frus- trate any systematic approach to the policy process. For instance, the his- toric Education Sector Review Program initiated during the early 1970s to bring fundamental educational reform in the country was aborted before implementation, having been politicized and polarized by opposing polit- ical agenda of the stakeholders, which eventually contributed to the fall of the imperial regime in 1974 (Country Studies, 1991). The formulation and implementation process of the current language policy is also a typical victim of this inadvertent trend. Soon after the current government took over as a transitional government in 1991, it convened a Conference for Peace and Democracy in Addis Ababa from July 2–6, 1991 (Ayalew, 1999). The conference, among other things, issued a policy guideline for the immediate provision of primary school instruction in five major ethnic languages. In addition to this, a decision was reached to allow choices of scripts in which the languages were to be written. Accordingly, the Latin alphabet replaced the Ethiopic alphabet for the Cushitic languages (which host most of the minority language groups) and the Ethiopic alphabet was retained for the Semitic language groups. This decision triggers some legitimate questions. First, whom did the conference include to make this major policy decision? Obviously, the conference was constituted of political parties that claimed to represent dif- ferent ethnic groups, but, as some scholars argue (for example Ayalew, 2000; Tekeste, 1996), there is no proof that the respective speakers of the language were consulted to check on their needs, nor was an attempt made to explore the pros and cons of the two scripts in terms of providing the needed linguistic and educational opportunities for children. What perhaps makes this decision even more politically motivated is the prejudice against Ethiopic alphabet because it is the script of the Amhara ethnic group that had been in power for over a century. From linguistic and economic points of view, one may argue not only the possibility of using the Ethiopic alpha- bet for the Cushitic languages, but also the likelihood of doing it at less cost. 154 Planning and Changing Instructional Language Policy in Ethiopia Involvement of major partners in the development of the policy. From its inception, the policy of language of instruction enjoyed the sup- port of top-level officials. As a result, it did not take much time to appear as a policy agenda, involving different partners in the course of its devel- opment. In retrospect, it seems worth inquiring as to the constituencies who were involved in formulating the official document (although it came long after several languages had been implemented). The task of formulating the General Education Policy (the lan- guage policy being one major component) was delegated to five sub-task- forces totaling about 42 members (Seyoum, 1996). Most participants in the taskforce were from the Ministry of Education, the Addis Ababa Universi- ty, and development ministries (such as Health, Agriculture, Science and Technology, etc.). Once the draft document was ready for review, the Min- istry of Education held several meetings with teachers in Addis Ababa and seven other regional cities. However, the sad thing, as some writers, for example Seyoum (1996), accounted, was that no input was incorporated in the final policy as a result of the sessions held with teachers. An assessment of the draft document against the comments that the Ministry of Education had claimed to transpire shows nothing substantive, except rubber-stamp- ing. Worse, at that point in time (and still to date), the Ethiopian Teachers’ Union had split into two opposing factions (one pro-EPRDF/government and the other independent). As a result of this conflict, no important input came mainly from the independent teachers’ union to help improve the draft document. Table 1 attempts to sketch the major partners by level of the policy process, for further recapitulation of their roles. Table 1 The Involvement of Major Actors in the Policy Process Constituency Issue Agenda Policy Policy Imple- Evaluation definition setting formulation adoption mentation Office of the ** * Prime Minister Ethnic-based ** ** political parties Teachers’ Union ** * (government- affiliated) Federal Ministry ** ** * of Education University * professors Representatives * of development ministries (continued) Vol. 37, No. 3&4, 2006, pp. 151–168 155 Alemu Tekleselassie Table 1 (continued) Constituency Issue Agenda Policy Policy Imple- Evaluation definition setting formulation adoption mentation Multilateral and * bilateral foreign agencies Regional Educa- * tional Bureaus Teachers * and school administrators Parents * ** High degree of involvement. *Low degree of involvement Table 1 illustrates some striking patterns in the policy process. The general trend is the apparent shift of roles as the policy moves from its inception to its implementation. At the initial stages, the issue was identi- fied and defined by ethnic-based political parties, as this was very consis- tent with the ideological orientation of the new government. It was, therefore, not a surprise for the agenda to be acted upon by the top authori- ties in the Prime Minister’s Office, who promote ethnic federalism and belong to EPRDF—the umbrella of ethnic-affiliated parties predominantly controlled by ethnic Tigraway. The formulation of the policy was soon del- egated to the federal Ministry of Education, which subsequently solicited some participation from development ministries in addition to its own management staff. The process of policy adoption intensively involved the federal Ministry of Education. Given the pervasiveness of the issue, how- ever, the process of adopting the policy also required the involvement of top-level officials by way of approving and ratifying the document. As it is often the case, the intermediaries could not avoid imple- menting the policy. The important issue remains whether they participated in the development of the policy or not. Whereas the table shows the exclusion of grassroots-level professionals, the implication of which is obvious, what tends also to be a grave omission is that of parents who pre- sumably have more stake than anybody else in such a tenuous issue as the language of instruction. Major actors and their power relationships. Despite cultural dif- ferences, there are conventional approaches to the policy process. It is, for example, logical to see policy adoption come before policy implementa- tion (Fowler, 2004). However, stimulated more by political lobby groups, the development of the language policy in Ethiopia did not follow logical stages. When the education and training policy (one major component of which is the language policy) was officially adopted in 1994, several lan- guages had already been under implementation as per the former decision by the Council of Representatives at the 1991 conference. Thus, the educa- tion policy did not come up with the language policy on its own. Rather, it 156 Planning and Changing Instructional Language Policy in Ethiopia attempted to rationalize the policy that was already under implementation by ostensibly emphasizing its pedagogical advantage: “Cognizant of the pedagogical advantages of the child learning in mother tongue and the rights of nations and nationalities to promote the use of their languages, primary education will be given in nationality languages” (Ministry of Education, 1994, p. 15). Power relationships between the major actors in the policy process. The investigators adopted the PRINCE System of Power Analysis modified by Fowler (2004) to sketch retrospectively the key actors and their power relationships in the process of the instructional language policy of Ethiopia. The term PRINCE is an allusion to the political handbook of Machiavelli which Coplin and O’Leary (as cited in Fowler, 2004) referenced to develop a system of power analysis. Fowler (2004) modified their work from a five level scale into a three level scale as shown in Table 2. The positions range from +3 (strongly supportive) to –3 (strongly opposed) where zero is neu- tral. The likelihood of the policy implementation according to the PRINCE model is calculated by dividing the sum total in favor of the policy (89) by the grand total (in favor of policy plus against policy) disregarding the signs (89 + 31), which is 89/120 or 74%. Table 2 PRINCE System of Power Analysis of Major Actors in Language Policy Actors Issue position Power Priority Total FOR: Ethnic-based political parties +3 x 3 x 3 27 Officials of the Ministry of Education +3 x 3 x 3 27 Teachers’ Union (government-affiliated) +3 x 3 x 3 27 The Media (government) +2 x 2 x 2 8 TOTAL 89 AGAINST: Multiethnic/cosmopolitan parties -3 x 1 x 3 -9 Teachers’ Unions (independent) -3 x 1 x 3 -9 The media (private press) -2 x 1 x 2 -4 Representatives of mixed communities -3 x 1 x 3 -9 TOTAL -31 Source: Authors’ direct account Table 2 explains the power relationships between the major actors in Ethiopia who are instrumental in the development and enforcement of the language policy. Not surprisingly, the actors in support of the policy Vol. 37, No. 3&4, 2006, pp. 151–168 157 Alemu Tekleselassie had enough power and influence (74%) to get the policy in place. The government set the pace for the balance of power. The actors in support of the policy mainly constituted what Fowler (2004, p. 155) calls ethnic-based “non-education interest groups” that had been encouraged through signifi- cant sources of power. For example, since the major media are entirely owned by the government, multiethnic parties were not able to disseminate their agenda as much as the ethnic parties did. In addition, as the authors of this article who at the time were teachers in Ethiopia observed, ethnic-based parties and the government-affiliated teacher union were provided with enough financial resources. By contrast, the actors against the policy were labeled by the ruling party as chauvinists and anti-government elements, deprived of their source of power, and discouraged through various penal- ties. For example, journalists of the private press, authorities of the inde- pendent teachers’ union (Prisoner of Conscience Released, 2002), and members of the cosmopolitan parties were harassed and finally imprisoned at the time the controversy surrounding the issue of language of instruction was at its peak in the country. This was basically against the values of responsible discourse where “less powerful actors should be genuinely free to take positions or choose courses of action without fear of negative reper- cussions” (Robinson as cited in Fowler, 2004, p. 49). Embedded are opposing value orientations governing the major actors. While the actors in favor of the policy were gripped with the pro- motion of multilingual cultures, ethnic rights, and equality, the actors against the policy process emphasized its negative implications on the national integrity. The latter groups also expressed their apprehension regarding the policy’s thoughtlessness to anticipate the human and materi- al resources needed to implement such a colossal endeavor as changing the language of instruction. Policy Implementation At least by default, one can learn that the premium placed on eth- nic politics seems to dictate why the implementation of the language poli- cy came prior to its formal declaration. Still, the implementation of the policy after its official adoption has special features worth addressing. One special feature in this regard was the abruptness of the process. Awakened by the official provision of the policy, several ethnic groups became part of the exercise. The approach toward implementing the poli- cy has been characterized by extreme rush (Ayalew, 1999; Wagaw, 1999). In other words, the regions did not go for a gradual approach to introduce the policy for fear that the delay would end up in reversing their rights. Accordingly, the implementation of the language policy started immedi- ately after 1994, by the translation of books from Amharic into other eth- nic languages. The translations were also carried out for all grade levels at the same time. The use of new script for the Cushitic language groups required teacher-training institutes to offer short-term training to teachers who could speak the language, yet who were unfamiliar with the new script. This was done side-by-side with removing qualified and experi- 158 Planning and Changing Instructional Language Policy in Ethiopia enced teachers and administrators who did not speak the local languages and replacing them with others “for political reasons and for their compat- ible ethnic origins,” thereby compromising standards (Wagaw, 1999, p. 85). In the Afar region (one of the language minority regions), for instance, individuals as under-qualified as seventh grade dropouts were assigned as district (woreda) supervisors and as primary teachers. In addi- tion, no sufficient material preparation was made by the time the start for implementation was declared. After a decade of first unofficial and then official implementation, the then national director of curriculum and instruction, Dereje (2001), confessed that “the newly adopted media of instruction have little or no literary stocks, such as dictionaries, glossaries and other printed literature” (p. 51). With its entire shortcomings the lan- guage policy has lingered now for over 12 years, with instruction being offered in at least 20 languages. Problems in Implementing the Language Policy The course of implementing the language policy was so hurried that little time or opportunity was left for sound planning. As a result, by the time the policy was put into place, several problems frustrated its implementation. This section explores three apparent problems. The first problem is the issue of mixed communities. The 1953 UNESCO recommendation states that “if mixed groups are unavoidable, instruction should be in the language which gives the least hardship to the bulk of the pupils, and special help should be given for those who do not speak the language of instruction” (UNESCO, 2003, p. 28). The settlement pattern of people in Ethiopia is such that mixed minority communities live side-by-side with the dominant ethnic groups, especially in urban and sub- urban areas. When the new policy was implemented in 1991 (as per the decision of political parties), no arrangement was made for children in these communities. Stirred by the problem witnessed in almost all parts of the country, the Ministry of Education issued a circular to all the regions (Ref No.15/1–94/19334/11) in November 1992 (Ayalew, 1999, p. 35). The circular requires regions to offer education to these communities in Amhar- ic, the national language, employing different alternatives such as shift sys- tems, or assigning different schools. This intervention of the government was, however, not well received in some regions until recently. As a result, parents pulled their children out of schools, and enrollment witnessed a dramatic drop in the areas settled by mixed communities (Hoben, 1995). The second problem that emerged in the course of implementing the language policy was the mandatory use of instructional language. Two distinct positions have emerged from parents since the language policy has been put in place in Ethiopia. The first group of parents includes those who are obligated to send their children to schools that cater for ethnic language instruction even when the children do not speak the language, the sole determinant being whether the children’s surnames coincide with the ethnic group in the area. Under the circumstances, the very concept of mother tongue instruction is displaced for ethnic language instruction, Vol. 37, No. 3&4, 2006, pp. 151–168 159 Alemu Tekleselassie violating both the pedagogical principles that suggest children learn better using their mother tongue, as well as the UN (1989) Convention on the Rights of the Childthat states children shall not be denied to use their own language. As Hoben (1995) further observes the situation, the only recourse for this group of parents was to vote with their feet; hence, the falling rate of enrollment. The second group of parents includes those whose real needs for language choice have been downplayed by political authorities. These are a special group of four ethnicities—the Welaita, Gamo, Goffa, and Dawro—in the southern parts of Ethiopia with similar, yet distinct lan- guages. Reducing the differences among these languages to the differ- ences of language dialect, the policymakers put them together and came up with a sort of ‘fabricated’ language called WOGAGODA. The new language was created by artificially coining the first two letters of each of the four languages. The policymakers made this decision basically because of the resource implication that providing instruction in each lan- guage would entail. Not unexpectedly, the language so created not only lacked anybody to claim it, but its implementation faced serious opposi- tion. As Ayalew (1999) accounts, the appeals of parents and community members were finally heeded, and the decision was reversed, but only after lives were lost and property was damaged. Viewed in terms of policy development, the failure of WOGAGODAepitomizes the weaknesses of policymakers, which Fullan (2000) conceives as failure to understand the multiple realities of people, the main stakeholders in implementing the change. Under the circumstances of WOGAGODA, the decision makers not only underestimated the local realities and value systems of the com- munity, but they also displaced the very goal of language of instruction— that is, the rights of people—for economic reasons. The third problem stems from the lack of grassroots capacity and readiness to exercise decentralization. It is a well-recognized fact that introducing several changes at the same time reduces the potential bene- fits of the policy endeavor (Fowler, 2000). Most policy initiatives in Ethiopia, however, do not escape this mistake, the reason being that a new government assuming power comes up with sweeping policy changes. The language policy introduced by the current government is just one among several reform initiatives the educational system experienced in the early 1990s. Asweeping decentralization of educational organization and management, the introduction of a new teachers’ career ladder, the adoption of new curriculum, and the development of a new system of financing public higher education are among the new policies that were adopted parallel to the language policy (Ministry of Education, 1994). Exploring the several changes that the educational system has undertaken since its inception is beyond the scope of this paper. It is worthwhile, how- ever, to address the role of decentralization, given its pervasive impact on the language policy. Decentralization bears fruits when it engages the will and capaci- ty of the intermediaries (Ayalew, 2000; Fowler, 2000). Owing to several changes simultaneously introduced with the language policy (and which 160 Planning and Changing