Forgetfulness without memory: reconstruction, landscape, and the politics of the everyday in post‐earthquake Gujarat, India

For many good reasons, after natural disasters it is common to work with ‘memory’ as part of a collective catharsis and a globalized humanitarian logic. Long-term anthropological research on the aftermath of the 2001 earthquake in Gujarat, however, also demonstrates the significance of forgetting in local practice. Immediately after the disaster, people vowed to abandon the sites of their loss, leave the ruins as monuments, and rebuild anew on safer ground. In time, though, life returned to the ruins as the terrible proximity of death receded, as memories and new salience were shaped by acts of reconstruction. The article explores some of the political and social factors that make this form of forgetting possible – or even necessary. Evidence of earlier earthquakes in the same region indicates that such ‘forgetting’ has an established history. Together, ethnographic and archival materials combine to cast doubt over the emphasis on ‘remembering’ as the only ‘memory solution’ to suffering.

2001 earthquake, how this change of heart came about: how thought inspired by sublime terror eventually lapsed. In conclusion, I will raise questions about the valorization of memory as a form of post-disaster humanitarianism and the idea that societies learn and store 'indigenous knowledge' about events such as earthquakes.
I had conducted research in Gujarat before the 2001 earthquake and initially my interest was, if possible, to help my friends. My first visit after the earthquake was some months later and ended before religious violence engulfed parts of the state in 2002 (Simpson 2006). Emotions were running high, tragedy hung heavily in the air, and shock ran through the lines of most conversation. The earthquake seemed to have come as a cruel surprise. Some people I had known from before had fled to distant relatives; others talked openly about depopulating the district forever. The most sanguine wanted to reconstruct what was there before on new, safer ground. In short, everyone wanted to get away from the sites of tragedy and loss. There was no possibility of return, not even a stray thought.
Later, as earthquake reconstruction became a research project for me, I became aware of previous earthquakes in the region. I also learned that the origin myth of the former kingdom tells of how the future ruler pierced a subterranean serpent with a metal stake in order to stop the creature from writhing and causing the earth to shake. Colonial archives and early post-independence newspapers revealed that thoughts about radical relocation and abandonment were not new. However, those earlier sites of destruction had eventually been resettled to the point that there was no visible association with earthquake destruction. My own research gradually turned into a longitudinal study as the compelling twists and turns of reconstruction made it difficult for me to take my eyes off the action. In time, the visceral and explicit expressions of wanting and needing to be elsewhere that I had encountered in the early years gave way to the rehabilitation of ruined urban areas and the return of people to places they had said they would never return to.
An event as significant and as inclusive as an earthquake creates many related and interwoven stories. My limited focus here is on how earthquakes have collapsed the town of Anjar in western India, and how there has been a repeated return of some of those who survived to make new life amongst the ruins. Each earthquake has been accompanied by momentous events, which one might assume would make these earthquakes memorable, or at least unforgettable: the 1819 earthquake coincided with the consolidation of British colonial rule; 1956 saw Jawaharlal Nehru, one of the great men of modern India, inaugurate a new town at a safe distance from the ruins in the belief that the old settlement should be abandoned forever; and 2001 brought about the incredible tragedy of a large goup of school children parading for the Republic being killed in the exact same spot as the previous collapses of the town. It also brought Narendra Modi, a Hindu nationalist politician, to power in the state. Modi went on to become the Prime Minister of India -in no small part owing to the way he presented post-earthquake reconstruction as a form of efficient governance.
After ten years of anthropological research on the aftermath of the 2001 earthquake, the ruins had again been re-inhabited. It was then brought home to me in a number of encounters that my friends had forgotten many of the things they had earnestly told me in the years just after the earthquake. Struck by this, I returned to my early field notes and discovered that the words people used to describe the earthquake in 2002 or 2003 had little or no obvious relationship to how they described it in 2010. This is not to say that they had simply forgotten things, as one might forget keys when leaving home. The process was more profound, linked I now think to the changing urban landscape and political zeitgeist as things were put back together. These slow but deeply significant ethnographic realizations helped me move closer to understanding why Anjar had been rebuilt despite strong initial doubts over the safety of the site. The result of this, however, is that I have decided to write much of this article's ethnographic sections in the abstract (rather than including named persons) so as to avoid the possibility of painfully confronting my friends and other research participants with their own falsely imagined futures. There are two exceptions: the words of a public figure, Dr Shyam Sundar, who campaigned valiantly for Anjar and for what he believed to be right, and my friend Bhushan, whose words have already been broadcast.
The story presented here is underpinned by a long-term ethnographic engagement in the region and a broader survey of relevant literature (Simpson 2011). The source material on the 1819 earthquake was written to support the colonial projectsometimes pointedly so. The materials I have used on the 1956 earthquake are from The Times of India, although I have also interviewed now-deceased district administrators who were influential in what happened then. My research on the 2001 earthquake centred on the district capital of Bhuj, but I made frequent trips to Anjar with friends, planners, and politicians. The high death toll, irregularities in land allocation, colourful protest movements, and innovative planning ideas were some of the factors that brought me repeatedly to the town. The story of each earthquake is thus assembled from different sources, each necessarily having a different character. My intention in combining the three earthquakes in the same narrative is to suggest that the processes I describe in detail for the 2001 earthquake have precedent and history. The lining up of disasters is not intended to stand as a historical account of two centuries punctuated by the uneven rhythm of catastrophe.
The main sections of the article are episodic, built around events that accompanied the three earthquakes. Then, I turn to examine in detail the ethnography of resettlement following the 2001 earthquake. First, however, I briefly review some of the literature on memory and forgetting to highlight and bring together two approaches to memory and social life from the anthropological literature: Paul Connerton's bold equation of memory and landscape and Veena Das's approach to understanding the power of transformation contained within everyday life. This union suggests that if memory and landscape are mutually constitutive, then the everyday changes within a physical and discursive landscape of reconstruction bring with them a transformation of what is remembered within the minutiae of social life. The high-resolution focus on the everyday, every day, also draws attention to the almost imperceptible movements within the logic of life and grounding ties as the world changed along with those who survived the earthquake.

Forgetting studies
There is no theory of forgetting in what follows; rather I show how the material and political conditions that allowed people to be repulsed by rubble have transformed over the years to make those spaces seem attractive once again. As Gastón Gordillo (2015) discusses, the remarkable messages contained in ruins and rubble are easy to overlook. In some ways, the material is commonplace, just one disaster among the many that has been uneasily forgotten. This has, however, happened not just once, but three times in less than two centuries! In other earthquake-prone parts of the world, notably Italy, sites Questions about memory have a long history, but it was perhaps Maurice Halbwachs who asked one of the first questions that has sustained modern memory studies: 'What would happen if all the members of my family disappeared? ' (1992: 73). His answer underpinned a sociological demonstration of the socially constructed nature of memory (also Halbwachs 1980). In different ways since, anthropologists have confronted memory and time in particular locations (Appadurai 1981;Bloch 1977) and have made more general theoretical interventions (Connerton 1989;Climo & Cattell 2002;Lambek & Antze 1996) showing that memory is not only a psychological or cognitive question, but also connected to rituals, commemorations, bodily movements, and the many other ways we relate to one another, including, importantly, space and landscape. Looking back on the rapid expansion of memory studies in anthropology, David Berliner (2005) cautions us to note how 'memory' became virtually indistinguishable from 'culture' , 'myth' , and 'history' as the analytical framing of the concept was over-extended through the Holocaust, war, collapse of socialism, and the rise of neoliberalism.
Anthropological approaches usually foreground forgetting as a social process in which people negotiate and create different forms of identity. Forgetting in an anthropological register emerges as a form of selection or refinement, an exercise of strategic will almost. Alternatively, forgetting is something that is imposed by states, especially in former communist and socialist countries (Berdahl 2010;Martinez 2018) in relation to 'traditional' practices such as shamanism (Buyandelger 2013). In other contexts, forms of cultural praxis may be made to appear 'backwards' , 'superstitious' , or 'primitive' by the authorized modernity of some greater power, such as organized religion or the state.
In contrast, in the South Asian literature, forgetting has been strongly related to the sedimentation of partition violence and the painful origins of postcolonial nationstates. In other words, forgetting comes with building. Consequently, there is an explicit connection between trauma and dislocation, on one hand, and nation building and construction, on the other. For Urvashi Butalia (2003), forgetting the physical and sexualized violence of partition gave rise to the possibility of society. Forgetting makes it possible to continue and become part of a postcolonial narrative, beyond silence. Influential, too, has been Veena Das's theorization of the 'everyday' as a way of living with critical and violent events (Das 2007;Das & Kleinman 2001). For Das, the boundaries between the ordinary and the eventful are drawn in terms of the failure of the grammar of the ordinary. By this, she means that what is put into question is how we ever learned what kind of object grief, or love, is.
To this we might add how we learned of the importance of the stability and continuance of place and landscape. For me, Das's work resonates with what I have seen in Gujarat, where daily routines and order within urban landscapes were, and are, such important parts of life (Simpson 2005). For Das, the eventful is not distinct from the ordinary, for daily life buries within itself violence and other catastrophes which provide force within everyday relationships. The ordinary is not the mundane, for through routine there is healing and the ability to tame great events. The thoughts and presence of violence thus become part of the everyday. Paul Connerton's work (2009) has attempted to explain the forgetfulness of modernity (see also Lefebvre 2014). For him, such forgetfulness can be attributed to the instability and uncertainties of contemporary life, the rapidity with which places change identity, and the collapse of fixed status hierarchies. The ancient 'art of memory' is thus intimately connected to landscape and place. Following Marx, Connerton claims people have been gradually separated from landscapes and terrain by the operation of capital, giving rise to forgetfulness. With the development of industrial production, suburbs and new forms of town planning divided communities, and the commute to work separated sites of production and consumption, splitting individual loyalties and places of intimacy. In Connerton's view, as in my experiences of Gujarat before the earthquake of 2001, bazaars, notable bungalows, and religious buildings were once enduring monuments that gave people their place and a narrative about time and belonging. Such monumental, palatial, or civic architectures were personal and collective reference points, first as construction projects that punctuated public life, then more subtly in the orientation of space and ways of living. Applying Veena Das's ethnographic lens, we can project these principles into the everyday life of urban design, vernacular sociology, and domestic space.
Connerton had European modernity in mind rather than the abrupt ruination of Anjar by an earthquake and the arrival of the sickly abundance of a neoliberal post-disaster boom. In some ways, however, the destruction of an earthquake is a provocative way of exploring his thesis, as such events demonstrably alienate people from familiar townscapes: structures that have appeared fixed are revealed, in collapse, to be fragile and temporary. As many anthropologists have shown, town-or cityscapes are contingent and in-the-making, layered and multiple, locations of both decay and accumulation (Smith 2018). Increasing the resolution in the space cleared by Connerton's grand theory, Joost Fontein (2018) reminds us that there are different ways of understanding the relation between memory and landscape, which include environmental determinism and social constructionism, and ethnographic approaches to these questions generally necessitate a course somewhere between the two.
In western India, the state and the political classes also play key roles in setting the terms of this relationship. After the earthquake of 2001, town plans reorientated the landscape, widening streets, carving up older neighbourhoods for vehicular access, and putting land to use in zones to rationalize and order the accumulated confusions of the centuries. Further alienation of people from 'memory landscapes' , evidenced in angry protest, occurred during the state-led reconstruction process as suburbs were introduced to 'modernize' Anjar and reduce population density. Many embraced new life in the suburbs, but that is a different story to the rehabilitation of ruined sites and the gradual return of people to places they had once said they would never return to.
In sum, forgetting plays a key role in the South Asian literature, associated with growth and the construction of new things, nations, and lives. What I take from Veena Das is the ethnographic transformation of the intimate and the capacity of daily life to bring within itself extraordinary events, while Paul Connerton provides a means to think through the structural effects of wholesale modernity brought about by the interventions of development banks, the state, and industrialization. In the aftermath of all three earthquakes, it is clear that reconstruction brought about something new and not just the replacement of what was there before. Therefore, those who survived these disasters then had to contend with new realities, which, at least in the case of the 1956 and 2001 earthquakes, were alienating and accompanied by the imposition of new urban plans. Within this framework, and with the stretched ethnographic present of the 2001 earthquake, micro-shifts of everyday life accrue in ways that change individual and collective subjectivities to the point of radical restructuring over timesuch that what people said in 2002 about wanting to flee forever is no longer either an accessible memory or a possible thought because changed conditions mean that world has irrevocably gone.
The kind of forgetting I have encountered resists representation. However, one way to illustrate these more generalized processes is to recount a story about my friend Bhushan. In early 2002, he told me about the moment of the earthquake. I recorded him and featured the interview in a broadcast radio programme. He had been in the bath at the time of the disaster. When water started to run violently back and forth, he thought he had gone mad and death was upon him. In films, he said, lives flash before those who are about to die. For Bhushan, it was not like that. He found himself concentrating on a blue plastic pen holder. The object had fallen to the floor and broken the day before. In the swirling waters of madness, he clung to the image of broken plastic. That he could make it whole again gave him reason and strength not to let madness overcome him. Out of all the vain and glorious possibilities, loves and dreams, it was the hope of repairing a piece of plastic that saved Bhushan.
Skip forward nine years and Bhushan had forgotten the extraordinary story of the blue pen holder which saved his life. At his wedding in 2011, I asked him if he still had it. He had absolutely no idea what I was talking about. I left it for that moment -he was busy getting married. Later, I gently pressed him -but he had no ready recollection. He had forgotten the object to which he had once owed everything. It was not so much that he had forgotten, but that it was no longer there. The account of an everyday and ordinary object taking on all the significance of life is as compelling to me now as it was then. While it had stayed with me and in my notes as an example of how the movement of the earth can shake priorities and reassign hope, Bhushan forgot, without memory.
Today, Bhushan speaks of the earthquake in very different terms; of course, he has not forgotten the event itself. He sees it as the jolt that propelled the district into modern times, the moment that cemented the power of the nationalist political party he supports, and as the time when his brother started to plan his emigration to Australia. There is no madness or fear; these have gone. The description is of a different order, de-personalized and couched in the terms of a broader politically led narrative about development that has been aggressively promoted in the region. Bhushan's blue pen holder, itself a metonym for the earthquake, is the form of forgetting without memory that I take Maurice Blanchot to have had in mind in the words of the epigraph of this article.
As I have said, this kind of forgetting defies clear description and I have to rely on the juxtaposition of the story being there and then not being there. The material, however, asks: how is the unforgettable forgotten? Just as with Bhushan's pen holder, there is a point at which erasure happens on a larger scale and once-unthinkable sites are resettled. This vanishing moment may well have a psychological or neurological basis, but as an anthropologist working ethnographically in Anjar, such forgetting is also evidently part of a social process linking landscape and memory.
In Gujarat, over the same time period, the language of politics has been transformed by those who took the despair of the disaster and the widespread violence of 2002 and turned anti-Muslim hatred and discrimination into mantras about development and a confident and chauvinistic nationalist project. The consequence of these dual transformations is that the discursive and physical frames of reference that people such as Bhushan have at their disposal have also altered almost beyond recognition since  In the immediate aftermath, there was a strong sense in the town that Anjar should be relocated as such a concentration of death could only mean the land was unsafe. Who would want to dwell amid the memories and ghosts of such tragedy? The hastily convened cross-party Town Earthquake Relief Committee identified suitable land and pressed the political leadership for a resolution to this effect. Nothing of that kind happened, however, and by 2003, despite repeated, vigorous protests, the government looked set to permit reconstruction on the site. 3 In the intervening two years, another citizens' organization, Group 2001, had come to the forefront of negotiations between the people and the state. Led by the charismatic Shyam Sundar, its protest marches, strongly worded letters, and headline-grabbing soundbites ensured that the condition of Anjar received much press coverage and that its plight was well known across the region and at the forefront of much public discussion.

2001: The most recent fall of Anjar
I first met Shyam Sundar in 2004, three years after the earthquake. As we talked in his home, in front of a large poster of a ferocious-looking South Indian god, there were constant interruptions from his busy private hospital below. According to him, the desperate condition of Anjar three years after the earthquake was due to corrupt administration. He thought that reconstruction could have been swift and straightforward, but had been hijacked by the development banks, who had seen the opportunity to financially ensnare and reform the state.
His concerns having gradually shifted, Sundar was now campaigning against the rebuilding of Anjar on the land to the east of the old town, a geography given orientation by a building called 'MacMurdo's Bungalow' . He believed that the ground was dangerous and unsuitable for human settlement, having repeatedly collapsed in earthquakes over the last two centuries. He thought the government had hidden test results showing the area was built on lake sediment which made for weak foundations and amplified the movement of the earth. He repeatedly said, 'They are planning a cemetery not a town' .
Anjar was then still strewn with pathetic objects which recalled both the earthquake and the causes of Sundar's anger. Aside from an infamous temporary market constructed from rusting shipping containers, there was the brute fact that half the town was flattened and the other half was not. This was all the more obvious as the area to the east of MacMurdo's Bungalow had been cleared of rubble and the land was conspicuously bare. Daubed onto the wall next to the dark blue 'protected monument' sign outside MacMurdo's Bungalow were the red marks of planners indicating that a 6-metre-wide road was to pass through the middle of the historic building, as the increasingly confident nationalist government looked set to perform acts of historical erasure in the name of post-earthquake reconstruction. At that time, Sundar's campaigns were very well supported: it was obvious to him and his followers that Anjar should not be reconstructed on the same spot.
Elsewhere, the earthquake uncovered material evidence and neglected memories of previous disaster. In Bhuj, the capital of the former Princely State, the bastion of the inner fortress shed its outer skin of roughly hewn stone, revealing an older, elegantly ornamented façade of marching elephants and other portentous signs. The bastions had been encased following their partial collapse during the earthquake in 1819. The forgotten layers of history and meaning vividly uncovered return us to the beginning of the nineteenth century, when the convulsions of the earth coincided with the arrival of British colonizers and an unorthodox regime change in the palaces of Bhuj.

1819: Colonialism, regime change, and providence
From the beginning of the nineteenth century, it was British policy to 'settle' western India. To this end, the colonial government attempted to determine the boundaries between the many states so that regimes of tax and treaty could be imposed. Colonial influence thus seeped gradually from Bombay into what were to become its provinces. Kutch signed its first treaties with the British in the early years of the nineteenth century (McLeod 2007). 4 This led, in part, to the application of military pressure on the district. Colonel East crossed westwards across the Rann with 4,000 troops and took Anjar. Once established in the town, the new political agent James MacMurdo built himself a fine bungalow on a rocky outcrop in the shadow of a fortified tower. Some say the bungalow was modelled on his family home in Dumfriesshire in Scotland. He decorated the interior with beautiful scenes from the Ramayana painted in the local style.
Even from colonial accounts, it is obvious that the British were not welcome. These sources tend to depict Rao Bharmul, then the ruler, as a figure deranged by lust and intoxicants. James Burnes, who was surgeon for the Bombay establishment in Bhuj in the 1820s, wrote of how the Rao, infatuated in this hatred of the British, never spoke of MacMurdo in anything but the terms of the grossest abuse. After a series of skirmishes and near skirmishes between Kutch troops and the British, the Marquis of Hastings, the Governor General of India, proclaimed Rao Bharmul a public enemy and declared war against him. The Rao was deposed, his infant son nominated by the lesser ruling families of Kutch as successor, and the colonial administration established a Residency in Bhuj. James Burnes wrote of these events when they were recent, raw, and had yet to have the sharp edges taken off them by apologists or those seeking to rectify the more obvious prejudices of the colonial gaze. He wrote: The tyranny and injustice of Rao Bharmuljee had scarcely been crushed, and a new and better order of things introduced through the means of the British government, when the hand of Providence seemed to join in depriving Cutch of some of the instruments of cruelty. A violent shock of an earthquake, attended with some extra-ordinary circumstance, levelled with the dust nearly all the walled towns in the country, and anticipated an intention, which had often been conceived, of dismantling some of these nests of discontent and treason. The desolation of which can scarcely be imagined (Burnes 1829: 66 In Anjar, MacMurdo's attention was attracted by the slight motion of his chair. Before he could ask, 'What is that?' his chair again moved. Realizing an earthquake was upon him, he ran for open ground. Initially, he struggled to comprehend, his 'mind was too deeply occupied with the awful and appalling spectacle of the face of nature in a state of excessive agitation to admit of other thoughts or impressions' (MacMurdo 1820: 92). He observed that 'half of the town, which is situated in low rocky ridges, suffered comparatively nothing; whilst the other half, upon a slope to a plain of springs and swamps, into which the town is drained was entirely overturned' (1820: 99-100).
In the aftermath of 1819, the elegant walls of the old regime were encased in roughly hewn stone -a project presumably intended to make the palaces impervious to nature. The young ruler grew up to head a tributary of the Bombay Presidency, and became a favourite of Queen Victoria (Rushbrook Williams 1958: 232). Under the 'stabilising influence' of Residency rule, even his disagreeable father became 'temperate … to a degree' (Burnes 1829: 72). In Anjar, an inscription was erected recording that during the minority of Maharaj Rao Daisul, the Regency ordered the reconstruction of the fort. Through this work, the inscription states, 'the subjects were made happy, and the city was rendered flourishing' .
In the aftermath, MacMurdo wrote: '[T]here does not exist even a tradition of an earthquake of any violence having occurred. The natives, therefore, were perfect strangers to such a phenomenon, and were terrified in proportion to their ignorance' (1820: 104). Some years later, the travel writer Marianne Postans (1839: 85) noted, despite the destruction of 1819, that earthquakes do not form part of any local legends or tradition, suggesting perhaps that forgetting already had a history.

1956: Jawaharlal Nehru and the inauguration of new Anjar
At 9.30 p.m. on 21 July 1956, an earthquake destroyed 1,353 houses in Anjar, killing 115 people. 5 Most deaths occurred on the slopes to the east of MacMurdo's Bungalow. As a sign of the times, in Bombay, a fund-raising 'Quake Ball' was held at the Taj Hotel with Mickey Correa, 'the Bombay Jazzman' . 6 Elsewhere in the swinging city, R.N. Joshi, the Chairman of the Indian Institute of Engineers, spoke publicly about how the same part of Anjar had been destroyed also in 1819. 7 Minister for Revenue Rasiklal Parikh boldly told the Bombay Legislative Assembly that the plans to rebuild Anjar included the provision of 'quake-proof' houses. 8 Prime Minister Nehru visited Kutch in mid-August, offering words of cheer and condolence. He asked: how did you count the number of houses destroyed? How did you remove the debris and excavate the dead? Have people who left property in their houses claimed their valuables back? 9 He addressed a large crowd in Bhuj, the district capital, and told them: This and other natural disasters should not dishearten people. They should attempt at utilising the occasion, good or bad, to bring about some good results. The deaths and destruction were a matter of great sorrow. However, the sufferers should rebuild the devastated areas themselves and in a better way. Make them better places to live in. 10 At the time, it was reasoned that the old town of Anjar was built on dangerous ground and the earthquake provided an opportunity to construct a well-planned and modern settlement in a safer location. Nehru laid the foundation stone for New Anjar. 11  again became Anjar's busiest bazaar: the subjects were once again made happy, and the city was rendered flourishing. It was here that 184 children parading through the streets waving the national flag with pride were crushed as the whole area collapsed with the shock of the 2001 earthquake, a tragic inheritance of past decisions.

Return to the ruins
The process of returning to the ruins after the 1819 and 1956 earthquakes lacks narrative. No one recorded what happened and no one I have met has been able to tell that story. If you resettled Anjar, the destruction at the beginning of the story does not sensibly lead to the conclusion that houses should be rebuilt in the same location. Somewhere along the way, you must change your mind quite dramatically and run counter to what you had so viscerally felt immediately after mass collapse. However, despite the absence of a record, the strengthening of the palace walls in nearby Bhuj suggests an early nineteenth-century version of the modern slogan 'build back better' . The efforts following the 1956 earthquake at building a new township with well-ordered streets and land zoning suggest that at the time architectural and planning modernities (along with 'quake-proof housing') were seen as fitting social and material solutions to an unstable earth.
Anjar had developed in fits and starts over the centuries. The town's form reflected interwoven histories of caste, class, prosperity, and poverty, as well as changing fashions and technologies of construction. In 2001, those charged with mending Anjar faced a monumental question: are we to act on our sympathy for the population and patch things up quickly without doing further damage? Or are we going to make things worse now so life in the town will be better in the future? The earthquake initially showed itself to be stronger than the state. There was no functional post-disaster policy and consequently no one came forward with clear notions about what should be done. To consider the options through conventional governmental arrangements would have taken years. As it was, conditions on the reconstruction loans offered by the development banks necessitated the privatization of key state functions such that normal procedures could be bypassed. The result of these combined circumstances? Uncertainty.
Within the question of whether things should be made worse nestled the urgent issue of where the town should be rebuilt. Whilst land was identified at some distance from the old site, the debate was emotional and complex; as with much of my research on the earthquake, the opinions and ideas I recorded in the early years of this project (2001 and 2002) were raw, untamed responses to tragedy -quite unrecognizable today even for those who were at the heart of the discussions. Time has changed their tunes.
Some of those I knew well in Anjar advocated the possibility of depopulating Kutch, abandoning the land to earthquakes and drought. Some did indeed leave for Mumbai and Baroda; a handful stayed away, but after a few months or years most returned. Some of those who stayed reasoned that the ruins should be left as a permanent reminder of the danger of earthquakes (as has been the case in other earthquake zones). For them, this was an opportunity to build a new spacious city, of well-designed streets, zoned industrial and commercial areas, and fabled 'earthquake-proof' buildings. In this new future, they could leave behind the haunting scenes of death and the lingering, malevolent spirits of those who died prematurely.
Within a week of the disaster, there was public discussion about damage to Anjar's landmarks, temples, and civic buildings. The collapsed structures of the town were powerful, emotive symbols through which to insinuate the destruction of a social order. Anjar had been a bazaar and market town, a bus intersection, a place of artisans and tradesmen, and a focal point for pastoralist communities. It had thus been a material representation of a socio-moral arrangement that brought different kinds of people together in particular ways. This had been found wanting by the forces of destruction and it was time for a rethink.
As the unwieldy machinery of post-earthquake urban governance began to find its footing, the people, the state, and the invisible policies of international financial institutions came face-to-face in the ruins. They often did not recognize one another because the government had been semi-privatized and the people had formed unorthodox coalitions in unusual circumstances. The levels of disruption, indecision, and anxiety inflicted upon earthquake survivors were, with hindsight, quite extraordinary. Buildings were razed, grounds levelled, the town was first redesigned in rumour, then on paper, and then again on paper and in rumour. For a few years, life was noisy, unsettling, and stressful. In the midst of all this confusion, people continued to get on with their daily business, waiting to see what was to happen to Anjar.
Most displaced people remained either with relatives or in the temporary shelters that had sprung up on the town's outskirts. On the whole, these were tough places with unwelcoming communal facilities and poor-quality temporary buildings that were either too hot or too cold -but nearly always cramped as they also served as stores for possessions salvaged from the rubble. There was talk of these sites becoming permanent suburbs, with ordered lines of new housing. There was also the possibility of new land for housing being allotted on the basis of property lost to the disaster, or of longerterm residence in a temporary shelter. In other words, decisions about housing were far from clear-cut. Compensation schemes, property disputes, and proposed new town plans tended to further confuse the picture. Importantly, the choice about whether to stay or go was one that was played out over years as the options changed.
At irregular intervals over ten years, I visited the same locations in Anjar, met friends, and spoke to as many other people as I could. During these visits, I would make time to sit on a veranda just down the slope from MacMurdo's Bungalow. The view over the land to the east of the bungalow took in the section of Anjar that had repeatedly fallen; it was the same view MacMurdo had had to watch rebuilding following the 1819 earthquake. It was not a huge area, perhaps 300 metres by a kilometre, but there was something powerful about this piece of land. Once on the balcony, it was difficult to look away. It was a space of contemplation for me and the others, mostly old men, who gathered there too to reflect on what they saw before them. Of the 1,578 people who died in Anjar town, 1,550 of them were thought to have died within our view. 12 I took photographs of the rubble, then of ground clearance, and a few years later of the appearance of the first buildings. Viewed together, these photographs chronicle the transformation from debris to the almost total resettlement of the area. After a year, the rubble was cleared and the monotony of the razed landscape was broken only by the many trees that survived both the earthquake and the ground clearance. The trees stood as reminders of a lost place: cups of tea in the shade, evening conversation.
Others who congregated on this balcony invariably asked me what I was doing. Was I a government surveyor? Was I planning the new town? When I spoke in Gujarati and said I was a researcher, then people would describe the buildings that had once stood on the now vacant land. Projecting their memories, as if onto a canvas, their hands Once the rubble had been cleared, those who had lived in the area would invite me off the balcony to listen to their stories unfold as we walked on the razed ground. There was nothing immediately recognizable to me because the bulldozers had removed all street lines and distinctions between buildings; however, my guides generally knew the way through the old streets despite the fact that there were no marks on the ground. Sometimes, they would cross-check their location with familiar landmarks in the distance to work out where they had lived. The colonial bungalow, shrines, temples, and the location of the main gate to their old town continued to provide orientation.
In those moments, we put ourselves at the centre of a world that was no more, triangulating the location of what had been with the aid of what remained. Walking on this land of compacted rubble and shards of former domesticity was never easy. Conversation was slow and deliberate, but the underlying sense of the place was always utterly familiar to my melancholic guides. 'We lived here and my brother lived over there' . 'This was the main bazaar area' . 'There was a marriage hall here and the newspaper seller used to sit under that tree' .
Once, in an unexpected shower of winter rain, I ran for cover as the earth turned to mud. I found myself in a doorway at the edge of the sorry site. A man who had had similar ideas about staying dry turned and said, 'The rain cannot wash everything away' . I have no idea what he actually meant, but it struck me as apt and profound.
From these tours of the Anjar that was no longer I began to get a sense of the social threads that bound people to place, of the things that made a neighbourhood, and the coiled, frustrated energies that would eventually force people from temporary shelters and their dreams of a new suburban house back to the sites of former destruction. It took a few years of government indecision and the return of other people to make them forget their post-earthquake ideas about turning their backs on Anjar. Drawn to drink tea on the balcony overlooking the area, these men would describe how certain places were associated with particular castes or religious groups, as well as with prescribed kinds of activity. It became obvious that knowing where people and things were in town was more than a matter of practical knowledge: it was a form of moral reasoning which gave predictable order and pattern to daily life -these things had made Anjar.
In this way, the vegetarian could hope to avoid areas associated with the pollution of meat products and meat-eaters. Those needing a padlock or a tractor tyre would have a pretty clear idea of where such things were sold. Of course, the patterns of urban life in pre-earthquake Anjar were not unchanging. They had, after all, been repeatedly rearranged by earthquakes, and the order of castes and substances throughout the town was far from neatly distributed. On the whole, however, the people on the balcony would speak about the town that was as if they knew everything there was to know; or, rather, as if what they knew was all they needed to know. Describing society in this way was a form of social currency; not knowing and appearing keen to learn (as an anthropologist might) was not the generally accepted paradigm of social intercourse or memory. There was little doubt in these conversations -even when shared across a group of people who might not always have known each other -that this was how things had been.
Thinking about these descriptions of a certain past, it was as if by limiting what they knew and ascribing predictable attributes to who they knew that people were able to disguise the inevitable and actual disorder of a town, which, by the time of the earthquake, had around 70,000 inhabitants (far too many for any individual to know more than a small selection). In sum, even if the town had actually been socially chaotic, the balcony conversations masked the disorderliness of the past. There the past was perfectly known, especially when compared to the chaos of the present, and order predominated, rather than the frisson and danger of disorder, which had been so efficiently brought to light by the earthquake.
In this nostalgic and moral view of old Anjar, the calls and prayers of religious buildings, the fixed times of running water, and the opening and closing of the bazaars structured lives and routines, firmly emplacing people in the townscape. They thrived amid the neighbourliness that came with living cheek by jowl with shared routines. Anjar had, in part, made them who they were. It was not only that they remembered themselves being and doing in the town but that Anjar had given pattern and texture to their reasoning. Living in the town had encouraged them to make associations between particular things and people. The town and their journeys within it had structured the ways they could think and imagine connections between people and places. It informed the senses of those who lived there, affording them certain scents and lights and honing their sensibilities and tastes in particular ways, but not in others.
As new road and pipe infrastructures were etched into the ground, the view began to change shape. The orientation of new roads took the eye in different directions, making new connections between parts of the emerging townscape. The layout introduced new and wider roads, but was carefully aligned around the location of some older sites, particularly temples and shrines. By 2004, private construction was evident on what Shyam Sundar had described as lake sediment. Confident reconstruction of temples, mosques, and shrines followed. The area had been home to the main bazaars and oncevaluable land. Shops began to appear along similar lines to before, houses coming later, in a much more haphazard manner. Today, empty spaces remain within the urban fabric, but an outsider could not tell that the town had been flattened by an earthquake less than twenty years before. It is as if the descriptions of my interlocutors are taking new shape, as the patterns of association in their sociology are being re-created on the levelled ground. Over the years, the view from the balcony has gradually disappeared, as the foreground and skyline have been filled in with concrete, and the society that had been made so obvious by its absence has been lost again in its becoming.
When I first met him, Shyam Sundar was angry with the government; in years to come, however, he became increasingly baffled by the people of Anjar, who, before his very eyes, were returning to live on the land they themselves had deemed uninhabitable. He maintained that the land was dangerous and an alternative site was needed. Friends said that because he was an outsider he could not understand local traditions, ways of being, and the pull of returning to the area. In their eyes, a sense of 'homeland' and routine drew them back to that difficult spot -as, too, did the desire for a return to the ways which had been broken first by the earthquake and then further by the indecision and alienation of town planning and reconstruction. Return was now an act of re-creation and healing -not a form of irrationality, as Shyam Sundar so publicly thought.
In the 1990s, I visited Nehru's New Anjar with a wedding party. Even then, without my eyes open to the spatial politics of post-disaster dwelling, I noticed the place was half inhabited, as if it had never gained traction. Then, as now, it had been easy enough to move people with the promise of new land and property but it proved harder to relocate centuries of routine, habit, and tradition. New Anjar had been designed in the spirit of Nehru's India: caste-free, secular, and with planned divisions of work/enterprise and home/domesticity. In the longer term, the town failed to keep people away from the charms of old Anjar. The new site lacked the communal and congregational spaces of caste and religion; it also lacked places of gendered sociality and public consumption. Artisans and shopkeepers, who did not necessarily divide their time and interests between worlds of home and work, did not set up production or retail outlets on the new site. The new town did not belong to the people of Anjar. The gift of a metropolitan government obsessed with modernizing India through planning and architecture, it brought subtle values and assumptions about the good life that were not shared in the region. People felt like strangers in their own environment. Simply put, the people may have moved but the traditions and ways of life so rooted in the old town remained where they had been -and eventually brought some back to them.
The Hindu nationalist Bharatiya Janata Party was in power at the time of the 2001 earthquake. The incumbent Chief Minister was replaced by Narendra Modi, whose government was somewhere between supine and complicit in 2002's statewide religious violence. A provincial language of anti-Muslim chauvinism sought to create a Hindu vote bank united by culture, history, and fear. The earthquake drew sharp attention to the importance of infrastructure and economic growth for populist politics. In the years that followed, tremendous and organized political will saw public language and zeitgeist in Gujarat grew increasingly bullish as 'double-digit growth' and 'prosperity' became buzzwords. With this upsurge of political confidence came the 'Gujarat Model' , in no small measure based on the infrastructural, fiscal, and industrial policies that characterized post-earthquake reconstruction. Gujarat became outwardlooking as Modi became the leader of India.
In the same period, Bhushan forgot his swirling bathwater and pen holder; the people of old Anjar forgot their fear, returning to rebuild their routines and traditions; and the language of the state and political classes shifted from hate politics to nation-leading development. Consequently, the intellectual and descriptive resources available to talk about the disaster have transformed almost beyond recognition. The things people said in the early years are simply unthinkable to them now. At first, I found myself protesting their denials, 'No, but you really did say that!' Later, I realized that reasoning with someone about a self they no longer knew was folly, if not to say irresponsible. Perhaps all long-term ethnographic research would reveal similar shifts in general structure and intimate subjectivities over time, but that these changes came about in the aftermath of a disaster when the emphasis is so often on the importance of memory as an agent of catharsis leads me to my conclusions.

Conclusion
Over the last few decades, there has been considerable investment linking memory and heritage to collective catharsis after natural disasters as part of a universal humanitarian logic. As Elke Weesjes, the editor of the Natural Hazards Observer, puts it at the front of a special issue on the subject: 'Perhaps more than before, societies remember disasters … Why do we remember and commemorate? And why is it important to have a wellrounded narrative of a disaster that includes both the emotional and physical impacts?' (2015: 3).
Echoing these sentiments, memory projects have been lavishly funded after many recent natural disasters. In Gujarat, international humanitarians funded a major memory and heritage project (Tyabji 2006), the focus of which was on royal traditions and ceremonies, with less attention to the ordinary (largely owing to the lack of material evidence). The overarching assumption was that this form of recollection would be beneficial for strengthening resilience and hastening recovery -and in many ways it was. Evidently, much good came out of collecting old photographs and stories.
However, through these types of work a universal humanitarian logic has romanticized memory as empowering social capital and good practice. Yet if for a moment we ignore the emotional tugs of ruination and the professionalized humanitarian impulse and look instead to the conventional wisdoms of anthropology, we see again that ideas about remembering, commemoration, and narrative-making are culturally and historically conditioned processes. In some places, these may be alien or inappropriate; in others, they might not gain traction for different reasons. The assumptions about therapy and healing contained within these ideas have also developed within particular medicalized and cultural contexts and are not universally applicable. Importantly, however, in situations such as the one I have described, memory is now an unwanted companion for those who were not offered a compelling suburban alternative to the realities of their own ground zero. Humanitarians have tied disaster preparedness and vulnerability campaigns to reconstruction programmes. Knowledge of what to do in a disaster is seen as a good method of improving the survival rate. It occurs to me that such campaigns are likely to meet with more success in areas where disasters are anticipated rather than where they have just occurred, given the need to forget so that one can get on with things, as I have described in this article.
The colonial writers referred to earlier were struck by the fact that this part of Gujarat had been tortured by earthquakes for centuries but there were no traditions or memories: there was no 'indigenous wisdom' . The foundation myth of the kingdom warns of earthquakes, but this does not seem to have a general mnemonic effect and is certainly not at the forefront of contemporary regional consciousness. Following the 1819 earthquake, the Regency attempted to out-construct nature by encasing the old fort walls with a robust new structure. This complacency was allowed to stand until 2001, when the elegant carving of the old wall was revealed. Some years later, it was covered up with badly laid stone and topped with kitsch and low-cost cast concrete decorations. In 1956, Nehru attempted to import a new type of settlement that in the longer term proved unsuited to the daily concerns of the local population. Following the 2001 earthquake, there was no attempt in Anjar at a wholesale state-led plan for a new kind of society; there were no strong campaigns to build earthquake awareness into the routine and fundamental fabric of life. In essence, Anjar was left to its own devices, and for many this meant getting on with things. In this spirit, I hope to have shown why the people of Anjar do not live with the ever-present threat of an earthquake: to live like that would be to live with madness and fear, and not to live at all. Those who returned to the old spaces did so not simply because of an innate and unthought attachment to the place, but because they could not leave when they wanted to, for when they said they were going to it was, for most, impossible.
Thinking ethnographically about why Anjar has been rebuilt repeatedly, we see that those who did choose to live in the new town of 1956 and new post-2001 suburbs were not able to break the ties of property, tradition, and community that bound them. Conditions were not there for them to make that break -however much they might have wished for this. Instead, they fell back to the old places and the healing work of familiar routines. They often described this in terms of a cultural symbiosis of people and place, being at home in a familiar landscape of temples and secular spaces. Taking a long-term view on the consequences of earthquakes shows that some aspects of memory, here at least, recede to the point of forgetfulness and that public attitudes towards reconstruction transform quite radically over time. This is a possibility that humanitarians who follow natural disasters could carry into their memory work by focusing on stability as a means of facilitating forgetting.
The longitudinal nature of this research shows that memory and forgetfulness are connected to the changing world around us -particularly to landscapes, but also to broader processes of political language and social life. Memories are not timeless, caught like flies in amber, but in constant transformative dialogue with the world around us, our memories in fact only belonging to us in part. The 'everyday' absorbs new materials, adopts new words, moves salience and beauty around, and conducts a host of magical operations which move us imperceptibly to become different people capable of very different constellations and scales of thought. As Anjar was rebuilt before our eyes, it became increasingly difficult to map old haunts onto the new townscape.
For Connerton (2012: 37), the type of forgetting entailed in the formation of a new identity is not so much the loss involved in being unable to retain certain things but rather reflects the gain that accrues to those who know how to discard memories that are of no practical purpose in the management of one's current identity project. Forgetting then becomes part of the process by which newly shared memories are constructed because the new set of memories is frequently accompanied by tacitly shared silences. The many small acts of forgetting facilitated by the routines and everyday reconstruction practices which these silences help to enable over time are not random but patterned. I would suggest that forgetting can also be a form of return to an older identity rather than the embrace of a new one, and that the rupture of the 2001 earthquake took some people away to the forgetful suburbs when that choice was the compelling one, while others had to choose to rediscover themselves by forgetting the terrible things that had happened to them.
Even as the return to old Anjar gained momentum, Shyam Sundar continued to voice his doubts. Engineers were perhaps now confident that new building codes would withstand future shocks. The government's soil tests perhaps really did say the land was safe. At least two teams of scientists have carried out studies in Anjar since the earthquake. The first concluded the 2001 event was one too many for the old town, and the difference in damage when compared with neighbouring sectors is not an indication of a higher intensity of earthquake impact (Chatelain, Guiller & Parvez 2008). According to them, buildings collapsed because they were old and poorly constructed, not because they stood on a vulnerable spot. According to another team (Rastogi et al. 2011), the patterns of seismic motion had most probably been determined by the 'site effect' of Anjar: the shape of the land and conditions of underlying soil. Hence, they concluded that besides the poor-quality buildings, 'site effects' (lake-like sediment) also contributed to the collapse of part of the town. Despite the uncertainty of conflicting expert evidence, rebuilding old Anjar continues. Either way, the next earthquake lies in wait as things are once again rendered flourishing and assume new beauty and relevance. Vulnerability has been reconstructed and the next disaster assured.
Memory and forgetting are not opposites or like supply and demand curves where more of one leads to less of the other. Amid the wholesale distress and trauma of living after the earthquake, there are competing forms of memory and forgetting, some obvious like memorials, rituals, and commemorations; others less visible: cracks in old buildings, realigned roads, and new street names. As the grounding ties and logic of life have regrown and rendered Anjar flourishing, a military plane breaks the sound barrier in the skies over the town, the noise reminiscent of an earthquake, and people rush from their houses and shops in fear.

NOTES
I am grateful to all who have steered me -directly and indirectly -during the course of this research. Initial fieldwork was funded by a Nuffield New Career Development Fellowship and the Economic and Social Research Council . For inspiration in Gujarat, I am grateful to Stuart Corbridge, Malathi de Alwis, and Dilip Vaidya. Niamh Collard and Fabio Gygi kindly read and commented on an earlier draft. I would like to thank Sara Schneiderman, Anthony Oliver-Smith, Vincanne Adams, Jennifer Kramer, Michele Gamburd, Chika Watanabe, Micah Hilt, Azhar Tyabji, and Jonathan Eaton for helping me sharpen the argument at the point of final submission. I would also like to acknowledge the Editor of this journal and anonymous peer reviewers for their significant and patient contributions.