We all have the tendency to interpret developments in the past more rationally than they in fact were. But historical processes that at first look “logical” and driven by necessity, at closer inspection, often seem significantly less straightforward. Even the most well thought out and carefully planned social actions are somewhat erratic if observed from nearby (Watter 1996, p. 22). Robert Merton raised an important aspect of this problem when he pointed out, in his famous article published in 1936, on the “Unanticipated Consequences of Purposive Action,” that actions don’t always lead to the objectives that have been set. We do not know enough to gain a picture of all their implications; we assume, incorrectly that actions that were successful in the past, will also be successful in the future; we allow short-term evaluations to weigh more heavily than long-term concerns; we let ourselves be led by basic values that incite us to take immediate action, even though we have not considered the consequences thoroughly enough; or we rely for our actions on a forecast of what will come, that influences the course of things, and by so doing, is self-defeating (Merton 1936, pp. 895–896).Footnote 1

While Merton’s analysis points to the reasons why actions do not lead to the desired result, sixty years later, Charles Tilly has raised the question of the other half of the problem: how is it possible that sometimes, despite all the unanticipated consequences, purposive social action leads, nevertheless, to the desired results? “All life,” Tilly claimed, “is filled with erroneous interactions, therefore with unanticipated consequences. But life also teems with error correction and responses, sometimes almost instantaneous, to unexpected outcomes” (Tilly 1996, p. 594). Such responses to error are, says Tilly, “neither instinctual nor random; they draw on historically accumulated shared understandings, from culture. They also take place within constraining webs of previously established social relations, and, in the process, alter those webs incrementally” (Tilly 1996, p. 598). It is the combination of Mertonian unexpected outcomes and Tillyan error corrections that gives us a proper understanding of social processes, and allows us to withstand the seductive power of “ex post facto rationalization” (Tilly 1996, p. 600).

The campaign: Phase I

In the present article, I build on the work of Merton and Tilly, using a stylized narrative about the British campaign to abolish the slave trade. This campaign started on 25 March 1807, when the “Act for the Abolition of the Slave Trade” received Royal Assent. My reconstruction stops at the end of the nineteenth century, when the British campaign had gained broad international support—support that was consolidated in 1919 with the establishment of the League of Nations and the International Labor Organization, two institutions meant to universalize the fight against trading slaves and slavery.

From 1 May 1807 onwards, the “Act for the Abolition of the Slave Trade” prohibited British citizens from participating in the purchase, sale, and transfer of African slaves—with the exception of ships that had left Britain before that date, and delivered slaves in the West Indies up to March 1, 1808. Just exactly what the motivation was for this Act is a matter of ongoing and passionate controversy, but it is clear that it definitely made the whole of the Atlantic slave trade illegal for British citizens, and, in that sense, it represents a real milestone in the history of human rights. This, in itself, should be both remembered, and studied critically.Footnote 2

The “Act for the Abolition of the Slave Trade” was the pinnacle of a 20-year campaign by abolitionists, probably “the world’s first social movement” (Tilly 2004, p. 33; Tilly 2005, p. 173, 199). Already, in the first months of it being established in 1787, the committee that had played a key role in this decided that, in its campaign in the years to come, they would not target slavery as such, but only the slave trade. At a later date, Thomas Clarkson, one of the co-founders of the committee, summarized the arguments behind this decision. Two considerations turn out to have been of overriding importance. In the first place, they assumed that after the abolition of the slave trade, slaves’ living standards would improve.

For, the planters and others being unable to procure more slaves from the coast of Africa, it would follow directly, whenever this great event should take place, that they must treat those better, whom they might then have. They must render marriage honourable among them. They must establish the union of one man with one wife. They must give the pregnant woman more indulgencies. They must pay more attention to the rearing of their offspring. They must work and punish the adults with less rigour (Clarkson 1808, vol. I, pp. 284–285).

In the second place:

They would not incur the objection that they were meddling with the property of the planters, and letting loose an irritated race of beings, who, in consequence of all the vices and infirmities, which a state of slavery entails upon those who undergo it, were unfit for their freedom. By asking the government of the country to do this, and this only, they were asking for that, which it had an indisputable right to do; namely, to regulate or abolish any of its branches of commerce; whereas it was doubtful, whether it could interfere with the management of the internal affairs of the colonies, or whether this was not wholly the province of the legislatures established there (Clarkson 1808, vol. I, pp. 286–287).

From the beginning, the British State realized that abolishing the slave trade would have no effect, if Britain were the only country to do so. The other European slave-trading nations would then simply take over the British share of the trade, so that, in the end, Britain would just be helping its competitors economically, to its own disadvantage. Therefore, it was of great importance to convince France, Spain, Portugal, the Netherlands, and other slave-trading nations to stop trading too. The negotiations to achieve this objective often proceeded very laboriously. Tactics such as offering enticements were tried, for instance in 1814, by offering France “a West Indian Island or a sum of money” in exchange for immediate abolition, or by offering Spain large subsidies if it would restrict its slave trade to the area south of the equator and abolish the trade completely in 5 years time (Nelson 1942, pp. 195–196).

The first symbolic success was booked when the Congress of Vienna of 1815 added an appendix to the agreement made there in which the undersigned agreed that the trade in “negro slaves” “has been seen by all just and enlightened men of all times as repugnant to principles of humanity and universal morals” (Berding 1974; Frowein 1994, p. 361). Although the practical implications of these words in Europe seemed to be limited, this declaration had a deep impact on the freedom fighters in Latin America. This partly explains why a number of anticolonial governments in Latin America abolished slavery and the slave trade between 1810 and 1830 (King 1944, pp. 388–389; Corwin, 1967). After Vienna, other agreements, the purpose of which were much the same, followed (Paris 1815, Aix-la-Chapelle 1818, Verona 1822), but these did not have much effect. The other European states mainly considered it as a pretext allowing Britain to dominate the Atlantic Ocean as a kind of international policeman (Coursier 1960, p. 442; Adams 1925, pp. 609–611).

There was no significant decline in the slave trade in the Atlantic region after 1807–8, and that is putting it mildly. The volume of the trade was temporarily reduced because of the Saint Domingue revolt (1791–1804) and the Napoleonic wars (1803–15), but then it revived again. During the years 1801–10, 683,000 slaves were transported across the Atlantic Ocean, during the years 1811–20 the number decreased to 599,600, but during the years 1821–30 it increased again to 694,400 (Lovejoy 2000, p. 146). At the same time, all sorts of reports reached Britain that seemed to indicate that the treatment of slaves in the Americas was getting worse. In the 1820s, veteran abolitionists published critical surveys intended to prove that slave mortality was extremely high, that slave families lived under very difficult conditions, and that religion was suffering damage. In particular, in the West Indies, the exploitation of the remaining slaves appeared to be on the increase. The price of cane sugar decreased between 1815 and 1834, after the Napoleonic Wars, which led the planters to try to maintain their profit margins by enlarging overall production. As a result, the quality of life for the slaves involved became even worse (Blackburn 1988, p. 421, 428). Despite the initial ineffectivity of the campaign, it was also rigorously continued in the years that followed. But a complementary campaign came into being at the same time, that had more radical objectives.

An escalation of objectives: toward a second type of purposive action

According to many abolitionists, the tactics agreed on in 1787 had obviously not worked well enough, and in their opinion, that was why it was now necessary to go a step further and work on the gradual abolition of slavery as such. From 1823 onwards, they set a new campaign in motion that appeared to be widely supported by the British public. Various factors contributed to the fact that, in 1833, an “Act for the Abolition of Slavery” was passed that, from 1 August 1834 onwards, forbade slavery in the West Indies, Cape of Good Hope, and Mauritius. One factor was that West Indian agriculture had slumped; sugar could be obtained easily from other parts of the world; and the Jamaican slave revolt of Christmas 1831 had made a great impression on the public. Trying to abolish slavery as such, marked the beginning of a new phase in the struggle. People no longer directed their efforts on the symptoms, but on the disease itself. The campaign rapidly developed its own dynamics. It spread to other parts of the Empire, especially to India (Banaji 1933, pp. 333–404), and it gained support from abolitionists in other countries, which became evident, for example, during the “World Anti-Slavery Convention” (London 1840), in which also American and French abolitionists took part (Maynard 1960). With this escalation, entirely new anticipated and unanticipated consequences arose, which, in turn, led to new error corrections. But that is another story.

The campaign: Phase II

Despite its initial difficulties, the campaign gradually gained impetus. In 1839, Pope Gregory XVI published his apostolistic letter In Supremo in which he declared the slave trade from then onwards to be sinful, even though slaves who had already been sold or who were born in the New World should not be released, providing they were treated mercifully (Quinns 2004; Priesching 2008). Gradually, more and more countries also joined the British campaign, at first through bilateral agreements, and later also through multilateral ones (see Table 1). Bribes often featured in these arrangements. Spain, for example, received payments amounting to £1,134,179 up to 1853 and, in that same year, Portugal had received a total sum of £2,850,965 (Lloyd 1949: 45).

Table 1 The abolition of the slave trade and right-of-search treaties by various states, 1802–1862
Full size table

In this, however, the problem arose that there was no supranational authority that could check whether the various states that had agreed to stop trading slaves were actually keeping their promises. This was why, in many of these agreements, a “right-to-search” clause was included—in fact, this was an extension into peace-time of war legislation—by which the contractual partners claimed the right to stop ships in international waters and search them for slaves.Footnote 3 Formally, every state that had signed a “right-to-search” agreement with Britain could themselves search British ships, but in practice, it was almost always the British who searched the ships of other nations, the British navy being much more powerful than the navies of other states (Fischer 1957, p. 72). Agreements of this kind were not only made with a large number of less important countries, but, in the course of time, also with all the important slave-trading nations—the last of which to sign was the United States, in 1862.

There were two kinds of right-to-search agreements. In the first kind of agreement, the crew of slave ships were tried by so-called Mixed Courts, composed of judges from both countries. Between 1819 and 1871, a number of such courts were established at Freetown, Luanda (Angola), the Cape of Good Hope, Surinam, Havana, New York, and other places. It is estimated that the courts condemned over 600 slave vessels and liberated nearly 80,000 slaves (Bethell 1966, p. 79; Emmer 1981; Martinez-Fernandez 1995). In the other kind of agreement, captured vessels were handed over to a tribunal of the country under whose flag they were captured (Adams 1925, pp. 611–612). Implementing the right-to-search meant that the Royal Navy stationed a permanent detachment off the coast of West Africa, with the task of seizing the slave ships of like-minded states.

Error correction 1

According to the first right-to-search treaties, British cruisers could only capture a ship, if there were actually slaves on board. This slowed the campaign down enormously, because, as noticed by a contemporary observer:

Under the defects of these treaties, the slave dealer could bring his vessel fully equipped with slave-decks, water-casks, and provisions; could land the merchandise required in his barter for slaves; could collect at leisure his complement of Negroes on the shore; could send out empty vessels to decoy away any cruiser in the offing; and could thus take measures for effecting security from chase and capture: so that all the vigilance and zeal of the preventive squadron was effectually baffled by the cunning of the slave traders (Smith 1848, pp. 6–7).

Moreover, under these conditions, slave ships were tempted to unload their illegal cargo quickly, somewhere on land, or even to throw slaves into the sea.

This was why, step by step, the conditions in the treaties were tightened up, first by also making ships that had actually had slaves on board liable, and later on, by including ships that had obviously had the intention to transport slaves under these conditions too. Indications for this last mentioned situation included the presence of shackles, bolts, or handcuffs, and also extraordinarily large supplies of drink and food, such as large numbers of water casks, an unusually large boiler, or an extraordinarily large amount of rice. An Equipment Clause such as this was included for the first time in a treaty with Spain, in 1835, and later in the Quintuple Treaty between Britain, Austria, Prussia, and Russia, among others (Lloyd 1949, pp. 46–47).

Error correction 2

Despite all these additional conditions, it turned out to be difficult to block the slave trade. So the idea of “legitimate trade,” first advanced during the late eighteenth century, was revived: if the African slave traders could be persuaded to develop morally acceptable and yet lucrative economic activities, then there would be no incentive any more for them to hunt and sell people. There were all sorts of products that could be grown in West Africa and sold for profit in Europe, such as palm oil, a product that was being used increasingly for making soap and lubricants. James McQueen published this idea in 1821, in A Geographical and Commercial View of Northern and Central Africa, but it took another two decades before the idea found some resonance. At the initiative of Thomas Fowell Buxton, and with the support of the British government, in 1841–42 the first attempt was made to put this idea into practice. An expedition along the river Niger would have to promote commercial agriculture in inland Africa by encouraging the local population morally and economically to grow cash crops for the European market. The expedition became a complete failure (Temperley 1991; Gallagher 1950). Many abolitionists continued to see legitimate trade as a proper method to fight the slave trade, however.

The set of incentives that the British introduced in the decades after 1807 became quite varied over the course of time. In the end, it contained all the variants that Charles and Chris Tilly (1998, pp. 74–75) have identified: commitment (ideological endorsement by other European and American powers), coercion (the right to search, punishment), and compensation (bribes and legitimate trade).

The campaign’s consequences

In the end, the campaign succeeded. During the period 1841–50, the number of slaves traded across the Atlantic amounted to 435,300, but by the period 1851–60, this had declined to only 179,100, and between 1861 and 1867, the number dropped even further to 52,600 (Lovejoy 2000, p. 146). The consequences of this success for the slaves who were already in the Americas appear to have been overwhelmingly positive. Indeed, just as the abolitionists had hoped, there is evidence that in many cases they were treated better. In the large majority of slave zones native slave populations began to experience endogenous growth (Klein 1999, p. 194). But there were also several unanticipated consequences.

Unanticipated Consequence 1

Insofar as the British interventions were ineffective, they led to the emergence of intense competition between slave traders and abolitionists. The more sophisticated the prosecution techniques became, the more they came up against equally ingenious smuggling methods. The turn-around times of ships on the African coast were shortened, to evade the patrols operating along the coast.

This led to the necessity of maintaining full-time representatives whose job it was to gather large groups of slaves over a long period of time so as to be ready for rapid loadings when the American slavers arrived. African traders or resident American business representatives now remained on the coast over many years and developed full-time facilities to accumulate slaves in large permanent pens so that their slave ships could be loaded rapidly and leave the African coast in days instead of the traditional months that had been the norm in previous years (Klein 1999, pp. 199–200).

The export was also decentralized. Faster slave ships that could traverse shallower waters were built for the purpose; slaves were disguised as members of the crew; new forms of risk-financing were introduced; and foreign flags were fraudulently hoisted on the slave ships to fool the abolitionists (Campbell 1989, p. 22; Klein 1999, p. 200).

Unanticipated Consequence 2

There are strong indications that, as the cross-border slave trade declined, the domestic slave trade in the Americas increased. In Cuba, the United States, and Brazil, slaves were transferred from declining to rising branches of the economy. In the United States, a significant slave trade between the Lower and the Upper South developed. The more settlers moved to the South and West to grow cotton and sugar, the more the flow of slaves increased. A similar tendency could be observed in Brazil and Cuba (Anon 1946, p. 92; Herbert Klein 1971a, p. 567; Klein 1999, p. 194).

Unanticipated Consequence 3

The reduced supply of slaves also had a perverse effect: the occurrence of the deliberate breeding of slaves. That there were US plantations where this happened was made very plausible by Richard Sutch (1975). Because, what other reason could there have been for having mainly female slaves living on such plantations, who each bore and raised a large number of children?Footnote 4

Unanticipated Consequence 4

It was only in some cases that the growth of the “legitimate trade” led to a decrease in the slave trade. In fact, “the end of the slave trade [...] did not cause the rise of ‘legitimate trade’. This happened simultaneously and independently; and the slave trade within Africa flourished long after the export trade had ended” (Curtin 1981, p. 96). In some cases, the transition to cash-crop production (palm oil, peanuts, rubber, etc.) gave an independent income to people of slave origin, but in other cases, it was the slaves who could no longer be sold to Europeans who were used to produce the cash crops (Hopkins 1973; Klein 1971b; Law 1995; Korieh 2000; Austin 2009). In Cameroon, for example, the internal demand for slaves increased, because the production of palm oil and palm seeds was extraordinarily labour-intensive (Wirz 1972, p. 63). Similar situations arose in many other parts of West Africa (Lynn 1997, p. 52). For the Bight of Biafra, there is evidence that the trade in slaves increased right up to the 1830s, in alignment with the growth of the palm-oil trade. Moreover, “The greatest increase in the palm-oil trade did not come as a result of Africans switching to it from the slave trade, but from British ships being freed for this trade by the prohibition of their participation in the slave trade after 1807 and the conclusion of the Napoleonic wars and the blockade” (Northrup 1976, pp. 358–359; also Williams 1973). As late as about 1900, at least a million and perhaps even as many as 2.5 million slaves lived in the Sokoto Caliphate (in Northern Nigeria). In comparison: in 1860, approximately 4 million slaves were living in the United States (Lovejoy 1979, 1981; Lovejoy and Hogendorn 1993, p. l).

Unanticipated Consequence 5

While the campaign reduced the transatlantic slave trade further (measured over the whole of the nineteenth century, the volume was 42% of the trade in the eighteenth century), the trade in slaves within Africa itself, and from Africa to Europe and Asia, not only remained intact, but increased (Lovejoy 2000, p. 142). It was typical that, even in the colony of Sierra Leone, Britain’s most important African base in the fight against slave trading, coastal slave traffic continued as before, because the demand for slaves from inland areas carried on, and even grew (Howard 2006). In the Ottoman Empire, during the nineteenth century, the demand for African slaves increased, partly because the Russian occupation of Georgia and the Caucasus brought about stagnation in the transport of slaves from those areas, and because the opening of the Suez Canal made transport from Africa much easier (Moore-Harell 1999, pp. 410–411).

In addition to the trans-Atlantic routes, there were three other great trading routes, the trans-Saharan trade, the Indian Ocean trade, and the trade from Ethiopia across the Red Sea (Table 2).

Table 2 Slave exports from Africa
Full size table

Unanticipated Consequence 6

In addition, it seems that especially along the western regions of Africa, slaves became also objects of conspicuous consumption. Rulers in Yorubaland, Congo, and elsewhere expressed their political authority, especially after 1850, by accumulating slaves as prestige objects (Graham 1965; Johnston 1988, p. 14; Kiernan 1972, p. 215). These sorts of slaves were sometimes offered up at funerals—this was, in fact, an older tradition, but previously there had been fewer victims to it. “The number, sex, and age of slaves killed at funerals of slave-holders depended on the wealth, influence and instructions of the deceased, their relations and social expectations” (Ojo 2005, p. 386). These ritual killings had the underlying effect of disciplining the other slaves by fear, particularly those who did plantation and similar work. As a result of diminishing exports, the number of slaves within Africa had increased. Their large numbers were a threat to their masters, especially because several recent rebellions in the Futa Jalon, the Niger Delta, and among the Yoruba had occurred (Collins and Burns 2007, p. 211). King Kwaku Dua I (1833–67) justified the practice to a missionary: “If I were to abolish human sacrifices, I should deprive myself of one of the most effectual means of keeping the people in subjection” (Collins and Burns 2007, p. 211, quoting Freeman 1844, p. 164).

Responses

The abolitionists at that time were unaware of the full extent of all these consequences. Nevertheless, it was certainly obvious to them that the measures that had been taken were insufficient to achieve their aim fully. Moreover, outside the Empire, the escalation strategy (the campaign against slavery as such) was still not successful enough. Therefore they felt obliged to develop additional strategies. Two kinds of response emerged.

Response 1

To expand the campaign to include East Africa and the Indian Ocean. Already since the capture of Mauritius in 1810, the British had attempted to curb the slave trade in the region, but for a long time hardly to any avail. The most important slave trade centers were the port city Kilwa and the island Zanzibar, both belonging to the Empire of Seyyid Said, the Imam of Muscat. The number of slaves exported from the region grew from roughly 6,500 in 1834 to between thirteen and fifteen thousand in the 1840s (Iliffe 1979, 42). Several agreements with Seyyid Said remained ineffective, although 1845 seemed to mark a turning point, when the sultan agreed to prohibit the slave trade within his empire, and allowed the British Navy to seize and condemn any slavers found in any seas as if they were British ships. But again, the agreement largely existed on paper. Only when Britain consolidated its rule in the Sudan and East Africa at the end of the nineteenth century, the Arab slave trade began to experience serious problems. In 1873 the sultan of Zanzibar was forced to abolish immediately his entire domestic coastal slave trade, to close down all public slave markets, and to make impossible all transit slave trade. Similar treaties were made with Egypt in 1877, the Ottoman Empire in 1880, and Persia in 1882 (Wilson 1950, pp. 520–521; Toledano 1982; Clarence-Smith 2006). But even then, the campaign was not over, since measures did not show durable effect. The Anglo-Egyptian Convention of 1877, for instance, seemed to signal a change because the Khedive Ismail was forced to appoint the Brit Charles Gordon as governor-general of the Sudan, and Gordon immediately began to suppress the slave trade. In 1879 Gordon’s troops defeated an alliance of slave traders, after which the trade decreased for a while. But a few years later the trade began to recover, since the local elite had no interest in total abolition (Moore-Harell 1998, 1999, pp. 417–420). Only at the beginning of the twentieth century the campaign´s success became clearly visible.

Response 2

A second reaction was to move the stage on which the struggle was being waged from the Atlantic Ocean to mainland Africa. If they could not succeed in ending the slave trade by taking action on the high seas, then they would have to go ashore and try to intervene directly, preferably in a peaceful manner (by convincing and by making treaties), but if necessary by force. The Niger expedition of 1841–42 had been a first attempt along these lines. Various methods were used in attempts to induce local African rulers to make the slave trade impossible. In this, they were not afraid of using force, nor did they hesitate in threatening to use it. Local rulers were put under great pressure and forced to sign treaties and accept heavy sanctions should they break these agreements. In the 1840s, more and more agreements were made with local rulers that included sanctioning clauses in addition to material incentives. For instance, it is stated in the “Treaty with the Kings and Chiefs of Bonny,” signed in 1844, that the leaders of Bonny (Bight of Biafra) agree that Great Britain was permitted to “put down the trade by force” and to carry out “severe acts of displeasure” against them and their subjects, if it turned out that the export of slaves had not been stopped completely and that the slave pens had not been burnt (Wilson 1950, p. 515). Acting on these kinds of “unequal” treaties, the British intervened, and did use force, as, for example, in 1840, when Commander Denman signed an agreement with the Chiefs of the River Gallinas, and “destroyed the property of the Spanish slave merchant there, rescuing and liberating the human chattels” (Wilson 1950, p. 514).

An additional influence on African affairs was the Christian mission. Encouraged by the British authorities (and, later on, also by the authorities of other “civilized” European countries) more and more Christian missionaries gradually began to establish themselves in West and Sub-Saharan Africa. Because they were dependent on supplies from overseas, they usually began by setting up stations along the coast. The Church Missionary Society, the Wesleyan Methodists, and the Baptists were amongst the earliest of these arrivals (Jakobsson 1972, pp. 62–130), but most of the church societies arrived during the 1840s, and thereafter. Helped by the introduction of quinine, that protected them from the previously fatal fevers (Headrick 1981, pp. 58–79), they dispersed rapidly into the hinterland, with the idea of not only spreading the Christian belief, but also of helping to replace trading in slaves with legitimate trading. For the missionaries, the religious and the economic aspects were often “two sides of the same coin,” as is evidenced by Livingstone’s slogan “Commerce and Christianity” and Buxton’s “The Bible and the Plough” (Johnson 1967, p. 177; Cole 1961; Stanley 1983).

Initially, the evidence is that missionaries were particularly successful with “the aliens, the foreigners, the uprooted” (Sundkler and Steed 2000, p. 88)—and among these were also many (ex-)slaves. Missionaries frequently offered shelter to people from these groups, finally establishing Christian villages where ex-slaves and escaped slaves could find refuge.Footnote 5 Father Joseph Lutz, an Alsatian member of the Congregation of the Holy Spirit, who worked in Onitsha, Nigeria from 1885 onwards,

founded Christian villages not unlike the Jesuit Reductions of 17th and 18th century Paraguay where liberated slaves, purchased from slave markets by the missionaries with charitable funds donated in Europe, could live Christian lives separate from the local people. Many of these liberated slaves, in any case, were not the local people but the hapless victims of slave-raiding further north, along the Niger and the Benue Rivers (Ryan 2006, p. 197).

Protestant missionaries often tried to do the same, although the management of these kinds of villages was sometimes rather dubious:

Since freed slaves had usually lost their tribal ties, this labour force was also somewhat movable; for example, workers were sent from Zanzibar to Masai, about one hundred miles inland from Lindi, by the Universities Mission to Central Africa, and to Freetown and Rabai near Mombasa by the Church Missionary Society (Johnson 1967, p. 178).

Both the treaties and the missions contributed to the fact that, in the second half of the nineteenth century, more and more land in Africa came under the direct control of the British. This is not to say that the work of the missions was used as a direct “instrument” for colonization (see, e.g., Porter 1993), but more that, in many regions “the missionaries obtained an ascendancy over the native populations, which could easily be turned into control exercised by political officials and magistrates” (Reinsch 1916, p. 45). Concurrently, however, the British came to the conclusion that, in their battle against the slave trade, they could not rely on treaties with the chiefs, because they kept breaking them, which meant that they had to be re-established again with policing expeditions. During several of these military campaigns, it became clear that the straightforward subjection of the local population was a more effective method.

The little forts on the West Coast of Africa, retained to suppress the slave trade, became the starting-points of a great inland expansion into Ashanti and the Northern territories, while Lagos, acquired in 1861 to put down the slave trade on the Oil Rivers, led to the expansion into Southern and Northern Nigeria by which 367,928 square miles were added to the Empire—a region more than three times the size of the United Kingdom (Knowles 1924, p. 78).

A third factor encouraged this movement: the turbulent economic growth in Britain itself, coupled with the ever-increasing demand for African raw materials (palm oil, etc.). There was a growing desire among British enterprises to control the production of raw materials locally, which meant penetrating and subordinating the African hinterlands to an ever greater extent (Wylie 1973, p. 215).

So we see that Chris and Charles Tilly’s three types of incentives all worked towards the same end in this situation: coercion (the enforcement of treaties), commitment (missionary work), and compensation (legitimate trade) all helped to further the colonial subjection of Africa. The consequence of this was that the slave trade dwindled more and more.

The Imperial British East Africa Company, formed in 1888, helped to bar the great slave route to Zanzibar on the east coast. France annexed between 1883 and 1896 Madagascar, which had been the great receiving depot for slaves, and thus put a serious obstacle in the way of the sea traffic in slaves. When the British took over Uganda and established a Protectorate in Nyasaland, they cut off two great sources of the supply of slaves. Kitchener conquered the Sudan in 1896, and put an end to the slave trade there, i.e. the southward and northward routes were both closed (Knowles 1924, p. 169).

Conclusion

In the end, the British campaign against trading in African slaves was largely successful, even though it has still not been entirely wiped out, even today. During the African conferences in Berlin (1884–85) and Brussels (1890), it was evident at last that all the important Western powers had reviewed their thinking on the slave trade and now wanted to give high priority to banning human trafficking. But the success of the campaign that, by then, had gone on for more than 80 years had caused side-effects of enormous proportions that no one in 1807 could have foreseen: the escalation of support for anti-slavery as such, and the takeover of a large part of the African continent by Britain and other powers. It is certainly conceivable, and even probable, that, even without the campaign against the slave trade, Africa would have been colonized. But the manner in which this colonization took place, and the way in which it could be legitimized by Britain and other European powers was, to a large extent, determined by the campaign. The General Act adopted in July 1890 by the Brussels Conference for the Abolition of the Slave Trade “cloaked the entire conquest of Africa in a humanitarian guise by presenting European rule and capitalist enterprise, including the employment of freed slaves, as anti-slavery measures. Thus, the ideology of the anti-slavery movement became part and parcel of the European mission to civilize Africa” (Roberts and Miers 1988, pp. 16–17).

The abolition of the slave trade underlines Charles Tilly’s (1996: 592) proposition that, despite unexpected outcomes, “purposive social action nonetheless produces systematic, durable social structure.” At the same time, my case study shows that the error corrections that were needed for this had extremely far-reaching consequences. The colonization of Africa was, to a certain extent, a second-degree unanticipated consequence of the campaign against the slave trade.

That the campaign had such far-reaching unanticipated consequences is probably due to the fact that it was all about an “absolute innovation” (Hadari 1989, p. 146). There had been other attempts at abolition, such as when the Dutch West India Company in 1738 decided not to take any more slaves to Surinam, or when Denmark forbade trading in slaves in 1803—but, in the first case, the motivation for abolition had been based purely on economic considerations (Jungman 2008), and in the second case, for Denmark, it only involved a minor European power (Johansen 1981). The British campaign, in contrast with these two earlier attempts, was not, in the first place, driven by economic motives and Britain was the “super power” of those days. In an underdetermined, and thus open, system, it is impossible to predict the consequences of a radical innovation such as the abolition of the slave trade. After all, “because we do not know ex ante which mechanisms will actually be at work (and perhaps have no knowledge of their mode of articulation) events are not deductively predictable” (Bhaskar 2008, p. 119; also Boudon 1985, pp. 165–190).

This case study suggests that three sorts of error correction can be distinguished, namely (a) measures that are taken to facilitate the effective implementation of the purposive action (e.g., the introduction of legitimate trade); (b) measures that are taken to minimalize unintended consequences, after the purposive action has been carried out (such as the agreements with local chiefs and the activities of the missionaries); and (c) new purposive actions that widen the objective and are more far-reaching than the original purposive action (the anti-slavery campaign).

Moreover, the case study indicates that the logic of error correction is not as simple as it sometimes appears to be (see Fig. 1). Karl Popper’s proposition for dealing with erroneous outcomes is to reduce the unnecessary insecurity of actions by not taking large steps, but by effecting the desired change “little by little” (Popper 1959, p. 75, 87); in that way, it should be possible to react faster and more adequately to unexpected outcomes, and easier to minimize unanticipated consequences. The British authorities seem to have done exactly what Popper advised, and yet the process went in a direction other than the one they had anticipated. An important, but dubious, assumption emerges in Popper’s argumentation. Imagine that a person (an actor) has set herself a certain objective. Within that framework, this “actor” has recently carried out action A1. The person now wants to proceed to action A2, to bring the objective closer. Piecemeal engineering assumes then, that before beginning on A2, the actor will know what the consequences of A1 are, because, only then can that actor make considered decisions about whether to repeat or intensify A1, or to amend and change A1 for another kind of action. It is possible, however, that not all the consequences of A1 will be evident until long after A2 has taken place—because information about it is not yet available; because the consequences are cumulative and so they could not all be identified at first; or because the consequences have been very late in appearing (Plé 1997, p. 183). In such cases, it is not evident why piecemeal engineering should be more effective than a more drastic intervention, based on Machiavelli’s (1994, p. 76) motto: “it is better to be headstrong than cautious.” Knowledge and error are reciprocally linked. “It is a paltry illusion to believe that they can be separated simply, clearly, neatly. It is an equally paltry illusion to believe that anything goes” (Castoriadis 1984, p. xiii.)

Fig. 1
figure 1

Purposive action, consequences, and corrections

Full size image