The English word compound, when referring to housing, possesses a surprising history. It has nothing to do with a being a mixture of elements, as in a chemical compound or a composition, but probably derives from the Malay word kampong, which means “village.”112 The word’s almost bucolic origins speak to the radical transformations to which the age of empire subjected it: compounds were what the British called the earliest colonial housing clusters in Southeast Asia—enclosures for European residences and factories. Later, the miserable, fenced-in dormitories where miners were housed on South Africa’s Rand were called compounds, and when larger employers in Lourenço Marques built worker housing, often big sheds to shelter hundreds of people under one roof, compound became componde.113 In the first decades of the twentieth century, conditions at these various dormitories in Lourenço Marques were considered scandalously abysmal, even by the low standards inherent to a system of forced labor.114 Neither authorities nor employers showed any sustained interest in financing a housing solution, either through paying higher wages or building livable homes. They were not eager, furthermore, to make permanent a workforce that it usually suited them to treat as transient.
In the 1950s, workers migrating to the city who did not have family to stay with and could not afford to build or rent a reed house of their own had no choice but to live in a compound. The typical compound featured a series of discrete units arranged in long rows on either side of a narrow yard.115 It was usually of wood-and-zinc construction (and therefore extremely hot), and rooms were often windowless and airless, with the door to the yard being the only opening. Packing people in so close together, it was more pigeon coop than barracks. The average room was less than 100 square feet, smaller than a room in the typical reed house.116 And people often tried to squeeze many to a room. At Araújo’s complex, four pit latrines served several hundred tenants, and the tenants were expected to maintain the facilities themselves. The landlord did give them free use of a water spigot in the yard, however.
Figure 1.18 A compound in Xipamanine, 1978. (Notícias archive)
Araújo made no secret of his business plan. It was the same as that of many cantina owners in the subúrbios: build a compound in the backyard to house sex workers who would cater to the bar’s clientele.117 But even though compounds were often stuck with the reputation of brothels, they were not brothels in the strict sense of being dedicated solely to prostitution.118 Not only sex workers lived there, and moreover, sex work was only one of a number of strategies that many of the compound’s young women and girls (and some boys), usually new to the city, were compelled to pursue. To live in a compound was to live in the slums of the subúrbios, and some homeowners and residents of longer standing in the bairros looked down their noses at their compound-dwelling neighbors. The especially rank conditions of most compounds contributed to the snobbery, as did ethnic chauvinism. To some native Maronga, the speakers of Chopi, Tonga, Tswa, and Changana arriving from farther north were unsophisticated country folk at best, unattached and potentially dangerous criminals at worst. That many of the so-called foreigners who lived in the compounds did so as a temporary strategy to accrue savings before returning to the countryside did little to alter the general perception of their rootlessness.
Until the late 1960s, rent in a compound could be very low if one shared a single unit with many others—some 100 to 150 escudos per unit. But when rents skyrocketed in Lourenço Marques and its subúrbios, compound living ceased to be the relative bargain it had once been.119 In 1971, rent for a single unit could be as high as 500 escudos per month if the compound had a water spigot and illumination. This was more than the rent for an average two-room reed house.
“The compounds exist,” argued Tempo magazine in 1972, “not for the benefit of the residents, who don’t even realize that it would be less harmful to live in houses of reeds—but rather because of attitudes dedicated to exploitation.”120
THE BAIRRO INDÍGENA
For years, the municipal government made repeated half gestures at the housing problem, always with meager results. In 1913, legislation compelling natives to register with authorities for the purposes of eventual labor impressment also mandated that the municipality dedicate a certain proportion of registration fees to the construction of formal housing for Africans.121 But over the next two decades, all that the municipality could show for its efforts was a cluster of thirty-three concrete-block houses near the market in Xipamanine, intended for “natives” who worked low-paying jobs for the municipality and the railroad. The houses lacked both piped water and electricity, though at least they were built in solid materials. The city charged a rent so high that few people with native status could afford to live there.122
In the mid-1930s, an additional source of funding for native housing was identified: indemnification funds resulting from Mozambicans who had died in the mines of South Africa.123 These funds had accrued for years, unspent, and the South African Chamber of Mines suggested to the Portuguese that they be used on something to benefit Mozambique’s African population. The governor-general of Mozambique revived the long-neglected order to build native housing, and he made the municipality of Lourenço Marques responsible for building it. The city was charged with building “a neighborhood that will come to serve as a model for others and to which can be transferred a part of the native population that currently lives, in the subúrbios of the city, in buildings of unpleasant aspect and devoid of the most basic hygienic conditions.”124
Figure 1.19 The Bairro Indígena, 1940s. (AHM, icon 42)
The project, which broke ground in the early 1940s, was called the Bairro Indígena da Munhuana (“the native neighborhood of Munhuana”). It was the colony’s first government-led housing development of any size intended for African residents. Other large, government-subsidized projects were under way by the 1960s, in the outlying areas of Matola and Machava and in other parts of Mozambique. But the Bairro Indígena was far more prominently located, and because it stood as the lone public housing intervention of any significance in Lourenço Marques for decades, it took on a symbolic value beyond the numbers it housed—for residents of the subúrbios, for colonial officials, and even for Frelimo both during and after independence. In the 1960s when the legal reforms of the time purged the term indígena from official communications, the name of the neighborhood was changed first to Bairro do Ultramar—ultramar (overseas) was how Portugal referred collectively to its territories in Africa and Asia—and then to the Bairro Popular da Munhuana. Yet even today, a half century after the name change, few call it anything other than the Bairro Indígena.
The 22-hectare site selected for the complex was located along a route that connected the City of Cement with the city’s airstrip, and if the project was indeed a superficial gesture—a Potemkin village only “for the English to see,” as the expression went—then it made sense to put it there, where many visiting VIPs entered the colonial capital.125 At the same time, the location was near some of the densest suburban neighborhoods.126 But there was a good reason the site was not so populous itself. It was a low-lying area that frequently was as inundated as the pestilential ponds that bordered it to the east and west. Reviewing