# Maji Maji: 75,000 or 300,000 Dead? Where Both Numbers Come From

Between 1905 and 1907 the German colonial army put down an uprising across the south of what is now Tanzania. The number of Africans who died is given, in almost every account, as „75,000 to 300,000.“ A range that wide is not a finding. It is two different findings standing next to each other, and they were made by different people, for different reasons, sixty years apart. Neither has been improved on since.

The low number is the perpetrator’s. The high number is a proportion. Both rest on the same unknown, which is how many people lived in the rebel area to begin with. This post is about that arithmetic, and about why the German conversation on colonial genocide has a Namibian chapter and almost no East African one.

What happened, briefly

The Maji Maji war began in July 1905 in the Matumbi hills, inland from Kilwa, as an attack on the cotton scheme: villagers were made to work communal plots for export, and were taxed by the head to force them into it. Fighters were ritually treated with water, maji, said to turn bullets to liquid. The revolt spread west and south through some twenty peoples who had no common political structure and, John Iliffe showed in 1967, only a thin organisational one. Several thousand fighters failed to take the German post at Mahenge on 30 August 1905. From October the Ngoni joined and were broken in turn.

The German answer, once reinforcements arrived, was not chiefly battle. It was hunger. In October 1905 Captain Curt von Wangenheim reported that only starvation would produce submission and that military action alone would amount to little. Troops burned villages, granaries and standing crops so thoroughly that they endangered their own supply. The governor, Gustav Adolf von Götzen, ordered that surrendering communities hand over their leaders and the „wizards“ who had spread the water; the records show these men were, as a rule, executed. The famine that followed lasted into 1908. Southern Usagara was described as emptied by 1906. A study made in Ulanga in the 1930s found that a quarter of the surviving women of childbearing age had become infertile.

German losses, by Susanne Kuss’s count, were fifteen Europeans, 389 askari and 66 porters.

The low number: a guess to parliament

Seventy-five thousand is the figure the colonial administration gave the Reichstag. Klaus Bachmann and Gerhard Kemp, in a 2021 article in Holocaust and Genocide Studies, trace it to the stenographic record of the 1907 to 1909 session, appendix 622, page 3693. It is, in other words, the German government’s own estimate of how many of its subjects it had killed, submitted to a parliament in which the Social Democrats were attacking colonial policy and the government had every reason to keep the figure low.

Iliffe put it plainly in the Journal of African History in 1967: „The official guess was that 75,000 Africans died.“ I quote it for one word. Guess. The man who wrote the standard history of Tanganyika did not call it an estimate or a count. Nobody counted. The administration had no census of the south before the war and none after; it had tax rolls, which shrank, and district officers‘ impressions, which varied.

So the floor of the range is not a floor. It is what the German colonial office was prepared to admit.

The high number: a third of a million

Three hundred thousand is Gilbert Clement Kamana Gwassa’s figure. Gwassa, a Tanzanian historian at Dar es Salaam, wrote the first major study of the war from the African side; his 1973 doctoral thesis was finally published in 2005. His conclusion, cited by Bachmann and Kemp, was 250,000 to 300,000 dead, which he expressed as roughly a third of the population of the uprising area.

That is where the population question comes in. Iliffe in 1967 wrote that perhaps a million people lived in the rebel area. „Perhaps a million“ is the base to which Gwassa’s third is applied. Tanja Bührer, in her 2011 history of the Schutztruppe, settled on 250,000 including deaths from disease and starvation. The one piece of district-level evidence Bachmann and Kemp cite points the same way: the officer at Tabora reported in 1908, on the basis of tax assessments, that a population he put at 750,000 to a million had fallen to about 500,000. That is a fall of a third to a half, in one district, measured by a method that also counts people who fled or stopped paying.

So the ceiling is a proportion times an estimate. Change the population base to 800,000 and Gwassa’s third is 265,000. Change it to 1.2 million and it is 400,000. The high number is at least as soft as the low one; it is simply soft in a more defensible direction, because it tries to count the famine and the low number does not.

Why the range has not narrowed

The two ends of the range come from two different questions. The German figure answers, at best, „how many died in the fighting and its immediate wake.“ Gwassa’s answers „how many fewer people were there afterwards.“ Famine deaths in 1907 and 1908, deaths from sleeping sickness that spread as tsetse followed game into depopulated land, children not born to the women of Ulanga: those belong to the second question and not the first. Nobody has done the work that would bring the two together, a district-by-district reconstruction from the tax rolls, mission registers and the 1913 colonial census, and the archives that would allow it sit in Dar es Salaam and Berlin, largely unread for this purpose.

There is also the matter of who is asking. Namibia has had, since the 1960s, Drechsler, Bley, Gewald, Krüger, Zimmerer, a Bundestag debate, a joint declaration and an official apology in 2021. Tanzania has had Gwassa, Iliffe, Becker and Beez, one 2021 legal article, and a presidential expression of shame on a visit in November 2023. The Herero and Nama had descendants in Germany’s public sphere, a lawsuit in New York, and a symbol, the extermination order, that fits on a slide. Maji Maji had twenty peoples, no single order, and a death toll that is mostly hunger, which photographs badly and litigates worse. The German colonial reckoning has been shaped by what could be made legible, and a famine across a region the size of Bavaria has never been made legible.

The genocide question

Bachmann and Kemp argue that the suppression meets the legal definition of genocide, and they are candid about what they did not find. There is no directive from Berlin ordering the destruction of any group. Götzen’s surrender conditions targeted leaders for having rebelled, not peoples for being peoples, and communities that stayed out of the rising were not attacked. In July 1906 his deputy relayed an order to apply the conditions flexibly, in the interest of preserving the population and its labour. The authors get to genocide by the route the Yugoslavia tribunal took at Srebrenica: intent inferred from conduct, the group defined as the perpetrators saw it, „the Blacks,“ and the killing of leaders treated as destruction of the group in part. That is an argument, and a serious one. It is not the same as an order, and the Namibian case has one.

Their Conversation piece, headlined with the 300,000, presents the top of the range as the toll. Their journal article, two years earlier, presents the range. Both are by the same authors. The difference is the audience.

What doesn’t add up

**A range whose two ends measure different things.** Seventy-five thousand is a wartime administrative admission to parliament, unsourced to any count and described by the field’s leading historian as a guess. Three hundred thousand is a fraction applied to a population nobody enumerated. Citing them as a range implies a single quantity with an uncertainty band. There is no such quantity; there are two definitions of death, one narrow and one wide, and the number for each is unknown. What would narrow it: a demographic reconstruction of the southern districts from tax records, mission baptism and burial registers, and the colonial censuses of 1913 and 1921, of the kind that has been done for the Herero and never for Maji Maji.

**The genocide case rests on inference, and the popular version drops the word.** The scholarly article says the intent is inferred and that Götzen’s own orders cut against destruction „in whole.“ The public version says 300,000 were killed and it was genocide. The archival gap between those two sentences is where the actual argument lives, and it is the part the reader never sees. What would settle it: not a number, since the legal definition does not turn on one, but the field orders and correspondence of the district commanders in the Tanzanian National Archives, series G, which Bachmann and Kemp cite for Tabora and which nobody has worked through for the southern districts where the famine was worst.

What stays open

How many people died in the south of German East Africa between 1905 and 1908 is not known to within a factor of four, and the reason is not that the evidence was destroyed. It is that the two sides who produced numbers wanted different things from them, and the third party who could reconcile them, a demographic historian with two years and access to two archives, has not yet been funded. Until then, „75,000 to 300,000“ will go on being printed as if it were a measurement, and every reader will silently pick the end that suits the argument they came with.

KardiKassin

Writes about Africa’s colonial history and how it gets told, sourced and checked at every step.

More from KardiKassin

Made in Germany: How Your Picture of Africa Was Manufactured

The same authors, two different numbers, depending on the audience. The free 12-page ebook traces that kind of drift across a whole continent’s worth of image-making.

Get the free ebook

Sources

• Bachmann, Klaus and Gerhard Kemp: Was Quashing the Maji-Maji Uprising Genocide? Holocaust and Genocide Studies 35 (2), 2021, pp. 235 to 249. https://doi.org/10.1093/hgs/dcab032 (open access; the source for the Reichstag figure, Gwassa, Bührer, Tabora and the Götzen orders)
• Bachmann, Klaus and Gerhard Kemp: 300,000 Tanzanians were killed by Germany during the Maji-Maji uprising. The Conversation, November 2023. https://theconversation.com/300-000-tanzanians-were-killed-by-germany-during-the-maji-maji-uprising-it-was-genocide-and-should-be-called-that-217712
• Iliffe, John: The Organization of the Maji Maji Rebellion. Journal of African History 8 (3), 1967, pp. 495 to 512; and A Modern History of Tanganyika. Cambridge University Press, 1979, ch. 6
• Gwassa, Gilbert Clement Kamana: The Outbreak and Development of the Maji Maji War 1905 to 1907. Cologne: Rüdiger Köppe, 2005 (doctoral thesis, University of Dar es Salaam, 1973), p. 217, cited via Bachmann and Kemp
• Stenographische Berichte über die Verhandlungen des Deutschen Reichstages, Anlage 622 (session 1907 to 1909), p. 3693, cited via Bachmann and Kemp