Key concept: change practice questions.

Watson and his colleagues studied the first two decades of McDonald's restaurants in Beijing, Hong Kong, Taipei, Seoul and Tokyo. They found that customers in several of these cities treated the restaurants as places to sit for hours over a single order, that children's birthday parties became central to the business, and that the company adjusted its premises and its staffing accordingly. Their conclusion about the chain's spread in East Asia is that

• local eating practices were left essentially untouched, since customers used the restaurants only as a novelty
• a standard product replaced local food cultures in each of the five cities in much the same way
• the transformation ran in one direction only, from the corporation's manual to the conduct of the customers who adopted it
• consumers reshaped what the restaurants were for, and the chain was itself altered by the settings it entered

Answer: consumers reshaped what the restaurants were for, and the chain was itself altered by the settings it entered

The case is the standard counterweight to accounts of globalisation as homogenisation: what arrived was taken up for local purposes, and the arrival changed the arriving institution too. Note that the argument is not that nothing changed — a new kind of public space for children did appear.

Yurchak, who grew up in Leningrad and later interviewed Soviet citizens of his own generation, described how by the 1970s the formulaic language of Party meetings was reproduced with great exactness by people who neither believed its claims nor opposed them, and who built ordinary lives in the spaces that this reproduction left open. He uses this to explain why the collapse of 1991 was experienced as

• both entirely unexpected at the time and, in retrospect, completely unsurprising to those who lived through it
• a change that most citizens had openly anticipated for at least a decade beforehand
• the work of a small number of dissidents who had rejected the official language from the outset
• evidence that the rituals had never been performed with any care by anyone who took part in them, least of all by the officials who led them

Answer: both entirely unexpected at the time and, in retrospect, completely unsurprising to those who lived through it

Yurchak's argument is that the form of authoritative discourse was reproduced ever more exactly while what it was for shifted underneath, so the system could look permanent and be hollow at once. The care taken over the performance is precisely his evidence, which is why reading the rituals as carelessly performed misreads the case.

Mitchell described a dance team on the Copperbelt of Northern Rhodesia in the early 1950s whose members wore carefully pressed European clothes, sang about life in the mining towns, and were identified by the rural district each of them came from. He found that these home-area labels mattered more in town, among strangers, than they had in the countryside. His conclusion is that such labels were

• a way of sorting strangers in a new urban setting rather than survivals carried in from the villages
• proof that migrants had kept their village arrangements intact through decades of wage labour
• a device introduced by the mining companies in order to keep the workforce divided
• categories in decline, as workers came to identify themselves by the job they held at the mine rather than by home

Answer: a way of sorting strangers in a new urban setting rather than survivals carried in from the villages

The finding reverses the expected direction of change: the categories were not remnants of rural life weakening in town but classifications that became useful precisely where people had to place strangers quickly.

Redfield's account of the Mexican village of Tepoztlán, published in 1930, described a well-integrated community with little conflict. Lewis restudied the same village in the 1940s and reported pervasive distrust, factional quarrels and disputes over land. Anthropologists usually draw from the pair the conclusion that

• the village must have changed beyond recognition in the years that separated the two studies
• the questions a fieldworker brings shape what a community is found to be
• restudies are unreliable, because the second researcher cannot observe what the first observed
• one of the two accounts must have been invented, since a single village cannot yield both

Answer: the questions a fieldworker brings shape what a community is found to be

Some change over fifteen years is certain, but it cannot account for a difference of this kind. The pair is taught because it makes the observer's framing visible, which is the argument for reporting the questions asked as well as the answers received.

Worsley's study of 1957 brought together colonial records of movements in Melanesia in which prophets announced the imminent arrival of ships or aircraft loaded with goods, and followers built landing strips and storehouses and left off ordinary work. His analysis presents these movements as

• episodes showing that the participants had misunderstood where manufactured goods are produced
• religious revivals with no bearing on the administration of the colonies concerned
• a consequence of the failure of colonial administrations to provide schooling in economics
• political responses to colonial inequality that organised people across groups previously separate

Answer: political responses to colonial inequality that organised people across groups previously separate

Worsley reads the movements as early political organisation under conditions in which open politics was closed off: a prophet could assemble a following across communities that had never acted together. Treating them as errors about manufacturing misses what they achieved.

Trouillot argued in 1995 that the revolution in Saint-Domingue, which ended with the independence of Haiti in 1804, was described at the time in terms that could not accommodate it, and was then left out of general histories of the period. He used the case to argue that

• the archives of the period were destroyed in the fighting, so the events of those years cannot now be reconstructed
• historical change is best studied through oral accounts, since written records are always partisan
• silences enter at every stage, from the making of records to the weighing of significance
• the participants themselves were unaware of the significance of what they were doing

Answer: silences enter at every stage, from the making of records to the weighing of significance

His point is that a silence is not simply a missing document. It is produced whenever an event is recorded, filed, retrieved or ranked in importance, so the absence has to be explained rather than treated as a gap in the evidence.