Making Africa Legible: Kiswahili Arabic and Orthographic Romanization in Colonial Zanzibar Caitlyn Bolton Abstract European colonialism and missionization in Africa initiated a mas- sive orthographic shift across the continent, as local languages that had been written for centuries in Arabic letters were forcibly re-writ- ten in Roman orthography through language standardization reforms and the introduction of colonial public schools. Using early mission- ary grammars promoting the “conversion of Africa from the East,” British colonial standardization policies and educational reforms, as well as petitions and newspaper editorials by the local Swahili- speaking community, I trace the story of the Romanization of Swahili in Zanzibar, the site chosen as the standard Swahili dialect. While the Romanization of African languages such as Swahili was part of a project of making Africa legible to Europeans during the colonial era, the resulting generation gap as children and parents read different letters made Africa more illegible to Africans themselves. Introduction Carved in wood amidst the swirls and arabesques typical of doors in Zanzibar is a phrase in Swahili: “There is nothing here but the house of God, and this is the door to heaven.” Yet the door in itself is hardly typical in Zanzibar, as it is the opening to one of East Africa’s first Christian churches, built in 1882 at the seat of missionization in the region. And neither is its writing typical, as the Swahili adorning the door to St. John’s Church was written in the Arabic script. Caitlyn Bolton is a PhD student in anthropology at the Graduate Center of the City University of New York, and an adjunct lecturer at Hunter College. Her research examines international Islamic development organizations’ interventions in religious education in Zanzibar. The Swahili language is an example of what Ali Mazrui calls the triple heritage of Africa, a continent indelibly formed by local African, Islamic, and European influences.1 Constructed primarily from both Bantu and Arabic lan- guages, written for centuries in the Arabic script, and rewritten in the Roman script during Christian missionization and British colonialism, Swahili bears the traces of these three influences and communities within its words, sounds, and shapes. The story of the Romanization of Swahili orthography in colonial Zanz- ibar is one of power, ideology, nostalgia, and loss at the nexus of these com- munities. It is set within the wider context of the massive orthographic shift in Africa, where indigenous languages spanning the breadth of the continent were forcibly re-orthographized from the Arabic to the Roman script during European colonialism – colonial powers even sought to write the Arabic of Sudan in the Roman script. As writing across the continent, from Wolof to Hausa to Swahili, shifted from Arabic to Roman letters, European colonial- ism was literally a project of making Africa more legible to Europeans. Ali Mazrui often told the larger stories of Africa, and in the case of Swahili his copious writings focus on its centrality in the political, religious, and eco- nomic spheres of the region spanning Kinshasa to Mombasa.2 Yet the story of the Romanization of Swahili orthography was a very personal one for him, as his father Shaykh Al-Amin b. Ali Mazrui was the chief voice opposing it in 1930s coastal Kenya.3 Drawing upon archival materials in English, Arabic, and Swahili from the Zanzibar National Archives, I tell here the local story of this orthographic shift in Zanzibar, the site chosen as the standard Swahili for the region: from the early missionaries’ first Swahili grammars and their at- tempts to “purify” Swahili of Arabic’s influence in their “conversion of Africa from the East,” to British colonial promotion of a standardized “literary” Swahili, to local Arab and Muslim opposition to Kiswahili Roman, and to the eventual colonial remorse as the orthographic shift facilitated an irreversible generational gap and social breakdown. While Romaniing Africa’s languages made Africa more legible to Europeans during the processes of missionization and colonization, the resulting generation gap as children and parents read different letters made Africa more illegible to Africans themselves. Language, Power, and Arabic Letters in Africa Language was always the companion of empire… language and empire began, increased, and flourished together. – Queen Isabel I of Castille (1451-1504)4 62 The American Journal of Islamic Social Sciences 33:3 “Romanization” here refers to the shift in writing African languages from the Arabic script to the Roman/Latin script, which constitutes the orthography of many modern European languages, including English. Yet Romanization has a deeper history that touches upon the enduring entanglement of language and power, as it refers historically to the expansion of Latin and the subsequent eradication of local languages that accompanied the spread of the Roman Em- pire. Other local languages may have continued to be spoken, but they disap- peared from the inscriptional record because “Latin defined writing and wrote the rest out of the record.”5 This orthographic conceit continued into the colonial era in the nineteenth and twentieth centuries, as the science of linguistics developed alongside im- perial rule and its contact with divergent languages and modes of writing. Early linguists pegged different orthographic systems along a developmental teleology that culminated in the Roman script, which was said to most clearly represent each phoneme with an individual character, while the Arabic script was, according to Johann Herder (d. 1803), “outstandingly imperfect.”6 In- formed by a belief in a primordial relationship between a language and the “spirit” of its people, early linguists also saw language and script as an index of the degree of “civilization” that a community exhibited along an evolution- ary trajectory that also culminated in Europe.7 Indeed, Eric Hobsbawm (d. 2012) calls philology, that early discipline of comparative and historical lin- guistics, “the first science which regarded evolution as its very core.”8 Since “civilization” in this context was not only linear and evolutionary but could also be sped up through “improvement” schemes enacted by those supposedly further along, Europe’s “civilizing” presence in Africa also en- tailed linguistic change, including writing down languages that had until then only been spoken and re-writing languages that had already been written in an “inferior” script. Linguists rose to provide support for these improvement schemes, including Richard Lepsius’ 1863 text, Standard alphabet for reduc- ing unwritten languages and foreign graphic systems to a uniform orthogra- phy in European letters, which detailed the proper method for writing Arabic in Roman letters and categorized Swahili as “illiterate” despite a centuries- long tradition of writing.9 While central to the ideological framework of improvement and “civi- lization” in colonial Africa, the orthographic Romanization of African lan- guages also signaled an immense shift in power relations on the continent. The ascendance of Roman orthography represented the “authority of colonial regimes over politically dangerous Islam.”10 In places such as Zanzibar, which had been ruled by Arab sultans, it also meant a literal authority over Arabs as Bolton: Making Africa Legible 63 Arabic was no longer, according to one colonial official, the “language of the ruling race.”11 In particular, it signaled authority over Arabic and Islamic regimes of knowledge: Arabic had been the language of learning and erudition in many parts of the continent, from Dakar to Dar Es Salaam, and colonial in- cursion entailed a shift not only in writing but in the entire educational insti- tutions that taught writing. Colonial officials called Qur’an school education “parrot-talk” and “dead- ening to the intellect,” and, while establishing their own schools, shifted the locus of authoritative knowledge and urged the acquisition of “useful knowl- edge,” often meaning that which might garner one a government post.12 Ro- manizing the Arabic orthographies of Swahili and other “Afro-Islamic languages,”13 such as Hausa and Wolof, changed not only the direction in which locals wrote their language, as students newly learned to write from left to right, but also changed the direction of entire communities. Sir Frederick Lugard (d. 1945), author of the British colonial policy of in- direct rule while acting as high commissioner for the Protectorate of Northern Nigeria and later governor of Nigeria as a whole, was a key figure in the Ro- manization of African orthographies – starting with Hausa. He instituted an active policy of “substituting the Roman for the Arabic character and of using Hausa instead of Arabic” as the language of government administration.14 De- spite his lack of knowledge of both languages, Lugard primarily justified the orthographic change with linguistic arguments regarding the unsuitability of Arabic letters to express Hausa phonemes. Yet he also expressed his concern for thwarting “Arabic Muhammadan progress” with the orthographic transition. Further, his colonial officers would be better able to monitor the corre- spondence of their African administrators.15 In indirect rule, while British colo- nial officers should have complete access to African correspondence, African administrators were to be barred from the language of their superiors: “I hope that, in course of time, this [policy] may result in the formation of a class of people who can read and write Hausa in the Roman character, though unable to speak English.”16 Later, as the chairman of the International Institute of African Languages and Cultures, Lugard aimed to lead the “‘clearing house’ for all information regarding the study and research in African languages and cultures.”17 In struggling against the Arabic script in Africa, he went so far as to ask his colleagues in the Anglo-Egyptian Sudan if they had yet succeeded in publishing Arabic books in the Roman script.18 As languages across West and East Africa were re-orthographized to Roman letters during the twentieth century, colonialism in Africa was literally a project of making Africa more legible to Europeans. 64 The American Journal of Islamic Social Sciences 33:3 Better Letters: Missionaries and the Early Codification of Swahili I have never regretted having rejected the Arabic mode of writing, which is too imperfect and too ambiguous for writing Suahili in a correct manner. – Ludwig Krapf, German missionary19 Swahili, known as Kiswahili to its speakers, has Bantu grammar and syntax and a vocabulary up to 44 percent of which is derived from Arabic.20 These Arabic loan words generally relate to those spheres in which interaction with Arabs were most frequent, namely, commerce and Islam, and, as such, over- whelmingly constitute the lexicon for trade, religion, jurisprudence, nonindige- nous flora, and maritime affairs.21 Given the centuries of trade via the Indian Ocean between East Africa and the Arabian Peninsula, Swahili was being written in Arabic orthography as far back as the eleventh century, as visible on coins and tombstones. However, the oldest currently available manuscripts date only from the eighteenth century.22 The first people to begin codifying Swahili, and in Roman letters, were missionaries who were preaching the Gospel on mainland East and Central Africa. Mazrui writes that in the early days of missionization in East Africa, Swahili’s connection to Islam was not held against the language, as some felt that this connection to a monotheistic religion set it apart as uniquely suited to convert Africans from pagan beliefs.23 Yet despite evidence that early missionaries did write Swahili in Arabic letters, such as on the door of St. John’s Church, that orthography proved to be an enduring problem for Christian missions in East Africa. It is in their early grammars that we find the first articulation of the unsuitability of Arabic letters for writing Swahili. The “imperfectness” of Arabic orthography was expressed first by linguistic arguments that thinly veiled issues of practicality for missionaries in learning a new language, and second, according to the belief that as the medium for God’s message, Swahili should be “purified” from any Islamic influence. Bishop Steere, head of the Universities’ Mission to Central Africa (UMCA) that was headquartered in Zanzibar as a base for missionizing the mainland, was a scholar of Swahili. In his 1884 Handbook of the Swahili Language, published by the Society for Promoting Christian Knowledge, he wrote “…it is impossible to adopt [Arabic letters] as the standard Swahili al- phabet,” despite the fact that this had already been the case for centuries. The reason “is that Swahili has five vowels, Arabic only three, and of Swahili Bolton: Making Africa Legible 65 consonants the Arabic supplies no means of writing ch, g, p, or v.” While he acknowledged that it would be “possible to express Swahili sounds by using [Arabic] letters with additional dots and affixing arbitrary sounds to the letters of prolongation, as is done in Persian,” he dismissed this option by saying that the “result would puzzle a genuine Swahili.”24 The Roman alphabet is “much clearer and better” he wrote25 and besides, as his German predecessor Ludwig Krapf (d. 1881) declared, “all the great societies of Europe and America had subscribed to that alphabet.”26 While Steere emphasized the difference between Arabic and Swahili, he minimized the difference between Swahili and English. He insisted that “there seems to be no difficulty in writing Swahili in Roman characters,”27 even though he recognized that the sounds in Swahili and English are not exactly the same. Despite this being a qualification for discarding Arabic let- ters, this was not a problem for Roman letters because it “generally happens that the sounds of a language, though not identical with those of another, cor- respond with them sufficiently to make the corresponding letters appropriate symbols,” such as the different pronunciation of letters between English and French, yet the French do not insist on writing the letter differently.28 Thus, he writes, “It is practically easier to learn to attach a new sound to a known letter, than to learn a new sound and a new letter too.”29 Clearly he is not speaking of the Swahili people themselves, who would have to learn not just one but an entirely new set of letters if Swahili were Romanized. Steere ad- vances his argument for the unsuitability of Arabic simply to facilitate the ease of evangelization in Swahili by European missionaries. As with Lugard in Nigeria, seemingly neutral linguistic arguments for an orthographic shift from Arabic to Roman letters mask the expediency of teaching the language to Europeans in a script that they already know. In addition to these linguistic arguments, missionaries opined that whereas Swahili was the prime medium for spreading God’s message in Africa, the use of Arabic letters and loan words would facilitate the spread of Islam in- stead. Steere saw his work as the “conversion of Africa from the East.”30 In- deed, with the five slaves he was gifted from the Arab sultan upon his arrival in Zanzibar in 1864, he began his missionary school and trained them to be- come the first African Christian missionaries to the mainland. They preached the Gospel in the Roman script. Despite Swahili’s Arabic and Islamic influ- ence, “through its negro structure, it is exactly fitted to serve as an interpreter of [the Christian] religion.”31 Evangelization did not necessarily have to be in one’s mother tongue, as with later Christian missionaries who insisted upon evangelizing in peoples’ “heart language.”32 Krapf writes that in the “great 66 The American Journal of Islamic Social Sciences 33:3 chain of mission stations which shall unite the East and West of Africa,” one need not translate more than “one or two Gospels as a beginning” before in- troducing Swahili as their primary “literary language,” because the number of African dialects “should be diminished.”33 While it was important to evan- gelize in a “negro” tongue, only one would suffice given the difficulty for missionaries to master them all. Missionaries were thus the most committed to “purifying” Swahili from its Arabic influence, again at times masked under seemingly neutral linguis- tic arguments. “In translations and grammatical works it is a rule of modern philology,” Krapf writes, “that all such foreign words should be rejected, and, wherever possible, only indigenous words should be used.”34 He con- siders “the best and most original dialect of Kisuahili” to be that of the Kenyan coast rather than of Zanzibar, because “the Kisuahili spoken at Zanzibar has a very large infusion of Arabic and other foreign words.”35 Later in the 1930s, Pastor Roehl took up this mantel and attempted to “re- Bantuize” and “purify Swahili”36 by attempting to translate the Bible in a Swahili stripped of its Arabic loan words. Instead, he drew upon the older Swahili words or those from other Bantu languages to fill the gap from the loss of the many Arabic religious terms. His attempt was judged a failure by a colleague, who counted 158 words of Arabic origin in Roehl’s first five chapters of Matthew alone.37 Further, the Bantu words that he adopted as more authentic carried pagan “ideas which one is particularly anxious to eradicate.”38 For example, “High Priest” was translated as Mtambikaji Mkuu, with mtambikaji referencing “an- imism” and denoting a medicine man who sacrifices to the ancestral spirits, rather than the common Arabic loan word “Kuhani,” which happens to also be a cognate to the original term in Hebrew (cohen).39 Later, when mission- aries and colonial officials sat alongside one another on the committee to stan- dardize Swahili, it was the missionaries who pushed to draw from English rather than Arabic when creating the new vocabulary needed to make Swahili a literary and administrative language.40 Discarded Letters: Arabic to Swahili in British Colonial Administration The suggestion to resuscitate the discarded use of Arabic characters for the writing of Swahili is absurd. It is thoroughly unsound from every point of view, and is universally condemned by all students of the language. – William Hendry, director of education, Zanzibar Protectorate, 1927 Bolton: Making Africa Legible 67 As Mazrui writes, the British in East Africa agonized for a while about the role of Swahili: “Should it be suppressed and discouraged as a rival to the English language? Or should it be promoted as something superior to ‘tribal vernaculars?’”41 As opposed to the Christian missions, the British colonial ad- ministration in Zanzibar did not at first utilize Swahili in its official commu- nications with residents of the island. During the time that Zanzibar was administered under the Foreign Office as a protectorate (1890-1914), colonial officials catered to the elite of the islands by publishing public documents only in English, Arabic, and Gujarati – perhaps assuming, as they later ac- cused local Arab elites of believing, that Swahili is a “slave language.”42 With Britain’s acquisition of mainland Tanganyika after World War I, the installment of Lugard’s protégé as Tanganyika’s governor tasked with implementing in- direct rule, and the transfer of Zanzibar to the Colonial Office in 1914, Zan- zibar’s colonial officials shifted their attitudes toward both Swahili and Arabic. They instituted educational reforms and teamed up with the other British East African territories to standardize Swahili and promote it as a literary and ad- ministrative language, chiefly promoting an ideology of language as having practical value above all else. The first public widespread utilization of Swahili by colonial officials was their campaign, during WWI, to “carry on a systematic propaganda” within Zanzibar to ensure support for the British during the war.43 In 1915, therefore, the colonial administration began publishing a weekly newspaper in both Ara- bic and Swahili called El Usbueyah (The Weekly), with Swahili in “both Roman and Arabic characters,” for “propaganda work and to enable the mass of the population to be acquainted with the latest telegrams concerning the war.” Further, the administration issued updates on the war on broadsheets in 1916 in Swahili written in Arabic characters, which they called “Kiswahili Arabic.” Clearly colonial officials did not share the missionaries’ assessment that altering Arabic letters to express the additional sounds in Swahili, as is done in Persian, would be confusing for Swahili readers, as this official pub- lication indicated the Swahili sounds of “p” and “v” with additional dots on the “b” and “f,” respectively.44 In the earlier years of British colonial admin- istration of Zanzibar, officials had no qualms about Swahili written in Arabic letters and even found a modified Arabic text suitable to express the unique Swahili sounds – as long as it allowed them to succeed in their propaganda efforts. This was all to change during the 1920s. With the acquisition of mainland Tanganyika from Germany after the war, Sir Donald Cameron, who had served as chief secretary under Lugard in Nigeria, was appointed governor of 68 The American Journal of Islamic Social Sciences 33:3 Tanganyika in 1925 to implement a policy of indirect rule. “It is our duty,” wrote Cameron soon after arriving in Tanganyika, “to do everything in our power to develop the native on lines which will not Westernize him and turn him into a bad imitation of a European," with the goal instead to “make of him a good African.”45 Cameron’s administration drew upon the former efforts of the German language policy in Tanganyika, which barred locals from learn- ing German and established Swahili – in Roman characters – as the language of administration.46 In the nearby islands of Zanzibar, leading up to the trans- feral of its administration from the Foreign to the Colonial Office in 1914, colonial officials began to see themselves not simply as “protecting” Zanzibar but actively influencing the population through policy and reform. These shifts in regional and local colonial policies were felt in the establishment of edu- cational programs and reforms along with participation in regional Swahili standardization efforts. Newly established public schools became the prime medium for teaching the public to write in the new Romanized Swahili. Local Qur’an schools had been conducted in Arabic and provided the only education on the islands until the twentieth century. In 1905 the colonial administration opened its first pub- lic school with an Egyptian teacher. While initial attendance rose, it dropped dramatically in 1910 when the administration decided to change the language of instruction from both Arabic and Swahili (written in Arabic script) to only Swahili written in the Roman script. The following year, this move was rec- ognized as a mistake because the “ground had not carefully been broken” and the “suspicion of the people” not yet overcome.47 In order to increase atten- dance, reforms in the 1920s reintroduced Arabic education, but only to the extent that it enabled people to read select chapters of the Qur’an that were explained via a Romanized Swahili translation. With the goal of eventually making education compulsory, the director of education wrote in a confiden- tial missive: “The truth is that Arabic is taught for no other reason than to please the Arab parents and to induce them to send their boys to school. Except in so far as it is required for religious instruction, it has no practical value to the protectorate.”48 Language, according to colonial officials, was taught according to its “practical value.” Arabic had no such value in the protectorate, and further- more, despite the fact that the islands were an Arab sultanate, it was deemed a “foreign language.” The director of education even wrote about “how im- possible it is to carry out any efficient teaching, if the first four years of a boy’s school life are complicated by learning a foreign language and two alphabets.”49 Arabic was no longer “the language of the ruling race,” they Bolton: Making Africa Legible 69 said, and the local Arabs’ “lethargy” in teaching their language, coupled with “concubinage with African women,” meant that Swahili is actually the language of the home: “When Arab children come to school it is rare even to find a child who knows the Arabic word for ‘mother’!” The “chief aim of the government in education” was to help “the young men of Zanzibar to be useful and loyal citizens,” and their policies around language education reflected that: dismissal of Arabic as a foreign and impractical language, and dismissal of Arabic orthography for Swahili, as learning Swahili in Roman characters might “be the means of obtaining a post in Government service.”50 Alongside these reforms in educational policy, colonial officials in Zan- zibar teamed up with their counterparts from Tanganyika, Kenya, and Uganda to standardize the language. Mazrui writes that “the very concept of standard Kiswahili”51 emerged with the acquisition of mainland Tanganyika because, with control over the entire territory, standardization actually became viable. The four territories formed the Inter-territorial Language Committee in 1932, tasked with standardizing Swahili vocabulary and orthography, and chose the language of Zanzibar as its model. “The policy of the East African Govern- ments is to develop Swahili as the common language of East Africa,” wrote Zanzibar’s Director of Education in 1938, and this “is the raison d’etre of the Inter-territorial Language Committee.”52 The committee translated famous works of English literature into Swahili, including Shakespeare, and held essay and other writing contests for local East Africans with monetary prizes, “half of which is to be spent in Swahili books.”53 They standardized spelling, to counter “the confusion which sometimes exists between f and v, p and b.”54 Yet such a “confusion” indicates that Swahili speakers may not have actually thought of the sounds as entirely distinct, given that only one of each pair is represented in standard Arabic orthography (“f” and “b”). This is not uncommon, as English speak- ers do not often consider as entirely distinct the two sounds associated with “th,” either that in “thank” or “the,” although these two sounds are expressed by separate letters in other languages, such as Arabic. In this case, even the colonial propaganda broadsheet from 1916 displays a “confusion” between these sounds, sometimes using the Arabic letter “f” and other times uses the modified “f” for “v” to represent Swahili words now written exclusively with a Roman “v” (e.g., vita [war] is written fita, while nguvu [power] is written as it is today pronounced, with the modified letter to express only “v”). 70 The American Journal of Islamic Social Sciences 33:3 Table 1: Arabic letters used to express more than one standardized Swahili sound55 Arabic Letter Transliteration Standard Swahili sounds expressed ب b b, bw, mb, mbw, p, mp, pw ت t t, tw ج j j, nj د d d, nd ز z z, nz ش sh sh, ch غ gh g, ng, ng’ ف f f, v, fy, vy ك k k, kw ي y y, ny And lastly, the committee oversaw the expansion of Swahili vocabulary to meet the growing needs of Swahili as an administrative and educational language, creating official terms for legal phrases such as without prejudice to the generality of and mathematical terms for denominator and fraction. On the question of from which language this new vocabulary should be de- rived, the missionaries on the committee were adamant that none should be derived from Arabic. Yet despite the ultimate decision that “each word should be considered on its own merits by the sub-committee,” the new terms that were drawn from Arabic were simply the repurposing of Arabic terms that had already been incorporated into Swahili. For example, denom- inator became asili, an Arabic loan word that already meant “origin” in Swahili. Fraction became sehemu, an Arabic loan word already meaning part. The entirely new words were all drawn from English, recast into “Swahili” spelling and pronunciation: Equator became ikweta, number be- came namba, and judge became jaji. This last term is significant, given the already wide proliferation of the Arabic term for judge, qāḍī spelled locally as kadhi. However, this term was associated with Islamic, rather than Euro- pean, jurisprudence.56 In evincing the ideology that language was chiefly for practical uses, and in this case the locals’ “useful” employment in the colonial administration, British colonial officials shifted their earlier support and employment of Ara- bic and Swahili in Arabic characters to Swahili in Roman characters only. By Bolton: Making Africa Legible 71 1927, the director of education had already referred to the use of Arabic letters for Swahili as “discarded” and “universally condemned.” Yet this condemna- tion was hardly universal, for the changes in educational and language policy were met with local outcry. Foreign Letters: Arab and Muslim Responses to Language Change Swahili is for the purpose of killing Arabic. – Al-Falak newspaper, The Arab Association, September 10, 1938 At the same time that Mazrui’s father was protesting the Romanization of Swahili orthography from the Kenyan coast,57 parents, religious leaders, and leaders in the Arab Association in Zanzibar were causing an uproar in the is- lands. “They teach Swahili in only foreign letters,” the prominent Arabs wrote in their petitions to the Arab sultan and the British resident, disturbed by the 1924 educational reforms. The Arab Association in Zanzibar Town followed with its own petition, writing that the new program “has greatly annoyed Mus- lims in general and Arabs in particular” and they “ask for total exclusion of Kiswahili in Roman Characters from the curriculum of the Government Schools.”58 The strident pushback from the local community, along with the poor quality of Arabic and Qur’an studies in the new schools, was so signifi- cant that the chief secretary later wrote that “the educational system very nearly collapsed” and “many schools were abandoned and have not yet been opened.”59 Swahili in Roman characters was a potent symbol that, according to one local newspaper, “the British policy of education had tried to smash the legacy of Arabic culture in Zanzibar.”60 This loss was important because lan- guage was not valued only for its practical use, but also for its capacity to transmit religious and moral teachings. The “curtailment of the Arabic lan- guage,” one local wrote in a newspaper, “is to a great extent responsible for the prevailing moral debasement amongst our young men and the corruption of manners and the want of Muslim-like qualities in them.”61 The new cur- riculum was an “onslaught on our language, manners, race and religion,” which “acts like a pickaxe in destroying mercilessly the Arabic language – the language of our religion.”62 Given the frequent declarations in the Qur’an that it is an “Arabic Qur’an”63 and thus the necessity for Muslims to pray in Arabic, the lack of facility in Arabic, even if it is not the language used at home, signaled a lack of facility in religion and access to religious moral 72 The American Journal of Islamic Social Sciences 33:3 teachings. One writer offered to gather for the government the “zunuj [black] citizens” to ask them if they would prefer learning Swahili over Arabic, and if they chose the former he would “break this pen and throw it to the sea.” He drew the following conclusion of the education reforms: “Swahili is for the purpose of killing Arabic.”64 The British colonial officials interpreted the pushback as due not to their own policies toward teaching Arabic and its letters, but rather to Arabs losing their prominence on the islands and their own lack of effort to preserve it. The director of education wrote that apprehension among the Arabs was due “not to the language and Islam being less efficiently taught than in former years, but to the fact that the language is in a dying condition in Zanzibar.” Further, he claimed that “Even far-seeing Arabs despise Swahili as a slave language.”65 Yet the local Arab and Muslim community had kept the lan- guage alive in the spheres in which it was valued (viz., education, religion, and literature) even if they did not speak to their mothers in Arabic. And while Swahili was not used in the same ways as Arabic, it could hardly have been seen as a “slave” language if it were their own mother tongue. Illegible Letters: Colonial Remorse and the Orthographic Generational Gap So long as Zanzibar owes its allegiance to Islam so long will the Arabic script live and be taught in Zanzibar. – F. B. Wilson, director of agriculture, 1939 After years of struggle to institute a new government school system that taught Swahili in Roman letters and a minimal course in Arabic, and after multiple reforms to try to assuage the local community’s concerns while also imple- menting their own agenda, the British colonial government ordered a review of the entire education sector. Their curious choice to head the commission, Director of Agriculture F. B. Wilson, is made less curious given the govern- ment’s focus on what they called “practical” education, namely, agricultural education, to encourage voluntary labor on the clove plantations in the context of the recent loss of slave labor.66 Wilson’s conclusion is that the “continuous struggle between the Director of Education and the representatives of the public” over education, as well as the abortive results of the government’s adult literacy programs, is due not to the apathy of a backward people, but rather to the government’s “refusal to use Arabic characters for Swahili writing and the stubborn insistence upon Roman characters which are generally disliked.”67 If his summary disavowal of the Bolton: Making Africa Legible 73 government’s language policy was surprising, even more surprising was his admission regarding the local Qur’an schools (katātīb/vyuo). While they had been repeatedly maligned in government reports as simply teaching “parrot- talk,” and confidential government reports indicate that the government hoped to phase them out with their government school program, Wilson writes that they actually do provide an adequate education in Arabic writing and teaching other virtues, such as respect for elders and an attitude of “self-help.” The result, especially in the rural areas, is that a large number of people deemed “illiterate” by government statistics were actually literate – but in Arabic letters. Writing Swahili in Arabic letters had not, in fact, been “discarded” as the director of education had claimed in 1927. In Wilson’s attached literacy report, he writes that due to religious training, “a very large number of people in Zanzibar, who are considered illiterate, are able to read and write in Arabic characters.”68 He found that across the islands, despite the transition to Swahili Roman in all public schools three decades earlier, 40 percent of the local pop- ulation was literate in the Arabic script, whereas only 2 to 3 percent were lit- erate in the Roman script. In an appendix he includes eight samples of “Letters Written in the Arabic Script by Zanzibar Peasants,” with the majority being letters written in Swahili Arabic with modified letters to express the sounds unique to the language. Yet above the locals’ dislike for Roman letters, a more pressing concern appears in Wilson’s report: the generational gap created by the new orthog- raphy. He writes of the “serious gap which is being widened by the teaching of the Roman script which the old people do not know,” where “young boys leaving school refer to their fathers as ‘wajinga’ (=ignoramuses) because they do not know ‘kiswahili cha kizungu’ (=the Roman script).”69 The younger generation learning Roman letters regards their education and language as “mainly utilitarian,” as a means of “qualifying for a Government post.” He looks down upon his father and, to the dismay of the director of agriculture, his father’s occupation as a cultivator. If only he learned his language in Arabic letters, he “would be brought up in the traditional religious lan- guage of their father with, I suggest, an atmosphere of greater co-operation and respect.” Wilson’s ultimate proposal is as shocking as it is simple: “The Arabic script should entirely replace the Roman script.”70 This was to apply to text- books, government notices, and “propaganda leaflets,” even to schools, which should teach Swahili in Arabic letters and possibly even arithmetic. Inserted into the copy of Wilson’s report in the Zanzibar National Archives is a note from the report’s editor to the central secretary suggesting that some of the 74 The American Journal of Islamic Social Sciences 33:3 redundant teachers could be usefully employed in transcribing textbooks and other materials in the Roman script into Arabic characters in anticipation of applying Wilson’s proposal of a massive orthographic shift back to Arabic let- ters. But that never occurred. While the following year’s annual education re- port mentions many of the other changes that Wilson suggested, it is remarkably silent on the subject of re-implementing the Arabic script. While the colonial government did not enact his proposed orthographic shift, one small result of Wilson’s literacy findings was that the government regularly posted a supplement to its local Swahili newspaper for the rural areas, called Kichumbe cha Maarifa. It was written in Kiswahili Arabic, with modified “b,” “f,” “gh,” and “j” Arabic letters to express the Swahili sounds “p,” “v,” “g,” and “ch,” respectively.71 Conclusion Given the enduring use of Swahili written in Arabic orthography throughout the colonial period, and further, the colonial government’s own employment of a modified Arabic orthography to write in Swahili, the early claims of missionaries that Arabic is “imperfect” and modifying its letters would “puz- zle the genuine Swahili” are proven baseless. Such claims, along with the orthographic project itself, were a product of ideology and convenience as missionaries and colonial officials sought to pursue their various agendas in the region while “civilizing” the population along European standards. Alongside Hausa, Wolof, and other African languages, the Romanization of Swahili orthography was part of a project to make Africa more legible to Europeans. Yet by their own account, as is recorded in Wilson’s reports, they failed. The prior system of schooling provided by the Qur’an schools and local ap- prenticeship better accomplished the educational goals of the British, namely, to encourage participation in agricultural cultivation. Arabic did not disappear, even though Arabs were no longer the “ruling race,” and its orthography is still taught today for Qur’anic recitation, even if Swahili is no longer written in Arabic letters. The local Qur’an schools did not die out as more and more students attended the government schools; rather, students began to attend both schools daily, a dual system that still remains today. But what they did succeed in doing was to create a generational gap where parents and children could no longer read each other’s writing. Africa may have become more leg- ible to Europeans alongside orthographic Romanization, but in the process it became more illegible to Africans themselves. Bolton: Making Africa Legible 75 Endnotes 1. Ali A. Mazrui, The Africans: A Triple Heritage (Boston: Little Brown & Co., 1986). 2. Ali A. Mazrui and Alamin M. Mazrui, Swahili State and Society: The Political Economy of an African Language (London: James Currey, 1995). 3. Ali Al’Amin Mazrui and Alamin Mazrui, The Power of Babel: Language & Governance in the African Experience (Oxford, UK and Chicago: University of Chicago Press, 1998), 73. 4. James Joseph Errington, Linguistics in a Colonial World: A Story of Language, Meaning, and Power (Malden, MA: Blackwell, 2008), 18. 5. Sheldon Pollock, The Language of the Gods in the World of Men: Sanskrit, Cul- ture, and Power in Premodern India (Berkeley: University of California Press, 2009), 271. 6. Errington, Linguistics in a Colonial World, 53. 7. Judith Irvine, “Mastering African Languages: The Politics of Linguistics in Nine- teenth-Century Senegal,” Social Analysis: The International Journal of Social and Cultural Practice 33 (1993): 27-28; Errington, Linguistics in a Colonial World, 55. 8. Eric Hobsbawm, The Age of Revolution: 1789-1848 (New York: Random House, 1962), 286. 9. Richard Lepsius, Standard Alphabet for Reducing Unwritten Languages and Foreign Graphic Systems to a Uniform Orthography in European Letters (Lon- don: Williams & Norgate, 1863). 10. Errington, Linguistics in a Colonial World, 142. 11. Zanzibar National Archives [ZNA] AB1/390. 12. Roman Loimeier, Between Social Skills and Marketable Skills: The Politics of Islamic Education in 20th Century Zanzibar. Islam in Africa, v. 10. (Leiden and Boston: Brill, 2009); Rudolph T. Ware, The Walking Qur’an: Islamic Education, Embodied Knowledge, and History in West Africa (Chapel Hill: The University of North Carolina Press, 2014); and Louis Brenner, Controlling Knowledge: Re- ligion, Power, and Schooling in a West African (Bloomington: Indiana Univer- sity Press, 2001). 13. Mazrui, The Power of Babel, 72. 14. John Edward Philips, Spurious Arabic: Hausa and Colonial Nigeria (Madison: African Studies Program, University of Wisconsin, 2000), 32. 15. Philips, Spurious Arabic, 33. 16. Ibid. 17. ZNA AB1/248. 18. Philips, Spurious Arabic, 34. 19. Ludwig Krapf, Outline of the Elements of the Kisuáheli Language, with Special Reference to the Kinika Dialect (Tübingen: 1850), ii. 76 The American Journal of Islamic Social Sciences 33:3 20. I. Bosha, The Influence of Arabic Language on KiSwahili with Trilingual Dic- tionary (Swahili-Arabic-English) (Dar-es-Salaam: Dar-es-Salaam University Press, 1993), 32. 21. Derek Nurse and Thomas Spear, The Swahili: Reconstructing the History and Language of an African Society, 800-1500 (Philadelphia: University of Penn- sylvania Press, 1985), 15. 22. Andrey Zhukov, “Old Swahili-Arabic Script and the Development of Swahili Literary Language,” Sudanic Africa 15 (2004): 1, 3. 23. Mazrui, Swahili State and Society, 52. 24. Edward Steere, A Handbook of the Swahili Language as Spoken at Zanzibar (London: Society for Promoting Christian Knowledge, 1884), 5-6. 25. Yahya Ali Omar and P. J. L. Frankl, “An Historical Review of the Arabic Ren- dering of Swahili Together with Proposals for the Development of a Swahili Writing System in Arabic Script (Based on the Swahili of Mombasa),” Journal of the Royal Asiatic Society 7, no. 1 (April 1997): 56. 26. Krapf, Outline of the Elements of the Kisuáheli Language, x. 27. Steere, A Handbook of the Swahili Language, 6. 28. Ibid., 7. 29. Ibid. 30. Cited in R. M. Heanley, A Memoir of Edward Steele (London: George Bell and Sons, 1888), 73. 31. Steere, A Handbook of the Swahili Language, iv. 32. Courtney Handman, “Speaking to the Soul: On Native Language and Authen- ticity in Papua New Guinea Bible Translation.” In Consequences of Contact: Language Ideologies and Sociocultural Transformation in Pacific Societies, ed. Miki Makihara and Bambi Schieffelin, 166-88 (Oxford: Oxford University Press, 2007). 33. Krapf, Outline of the Elements of the Kisuáheli Language, xi. 34. Ibid. 35. Ibid. 36. K. Roehl, “The Linguistic Situation in East Africa,” Journal of the International African Institute 3, no. 2 (April 1930): 199-200. 37. G. W. Broomfield, “The Re-Bantuization of the Swahili Language,” Journal of the International African Institute 4, no. 1 (January 1931): 80. 38. Broomfield, Re-Bantuization, 83. 39. Ibid. 40. ZNA AB1/239. 41. Mazrui, Swahili State and Society, 12. 42. ZNA AB1/390. 43. ZNA AB5/37. 44. Ibid. 45. Cited in Mahmood Mamdani, Citizen and Subject: Contemporary Africa and the Legacy of Late Colonialism (Princeton: Princeton University Press, 1996), 80. Bolton: Making Africa Legible 77 46. Marica Wright, “Local Roots of Policy in German East Africa,” The Journal of African History 9, no. 4 (1968): 623, 626. 47. Cited in Loimeier, Between Social Skills and Marketable Skills (Leiden and Boston: Brill, 2009), 290-91. 48. ZNA AB1/390. 49. Ibid. 50. Ibid. 51. Mazrui, Swahili State and Society, 44. 52. ZNA AB1/390. 53. ZNA AB1/239. 54. Ibid. 55. Cited in John M. Mugane, The Story of Swahili (Athens, OH: Ohio University Press, 2015), 184. 56. ZNA AB1/239. 57. Mazrui, Powel of Babel, 73. 58. ZNA AB1/390. 59. Ibid. 60. Loimeier, Between Social Skills and Marketable Skills, 321. 61. Al-Falak, July 23, 1933. 62. Ibid. 63. Qur’an 12:2, 20:113, 39:28, 41:3, and 42:7. 64. Al-Falak, September 10, 1938. 65. ZNA AB1/390. 66. This policy was influenced by the exchange of educational theory and practice with the American South beginning in the 1920s, particularly the Hampton- Tuskegee approach of “industrial education” for post-emancipation African Americans. See Kenneth King, Pan-Africanism and Education: A Study of Race Philanthropy and Education in the Southern States of America and East Africa (Oxford: Clarendon Press, 1971). 67. ZNA BA6/3. 68. ZNA AB1/390. 69. ZNA BA6/4. 70. Ibid. 71. ZNA AB1/246. 78 The American Journal of Islamic Social Sciences 33:3