After the 1989 Loma Prieta earthquake in the San Francisco Bay Area, power failures left food thawing in refrigerators. Rebecca Solnit later heard of people bringing it into the street and barbecuing with neighbors. An urgent practical nuisance became an occasion for cooking together and looking after others.
This cannot compensate for earthquake losses. It nevertheless drew Solnit toward behavior unlike the familiar expectation: in danger, people may quickly begin helping rather than simply fleeing. Disaster destroys much without automatically destroying the capacity to cooperate.
A Paradise Built in Hell, published in 2009, looks back through five disasters: San Francisco's 1906 earthquake, the 1917 Halifax explosion, Mexico City's 1985 earthquake, the September 11 attacks in New York, and Hurricane Katrina in 2005. Its paradise names relationships of mutual aid that briefly arose within catastrophe, not catastrophe as something desirable.
Combining survivors' accounts, historical records, and disaster research, Solnit asks why uncontrolled crowds dominate expectations if cooperation repeatedly occurs. What do those expectations then do to rescue?
A Crowd May Have Come to Help
Disaster films readily move from vanished order to looting, flight, and trampling. Such conflict is legible onscreen. Treating it as inevitable reality, however, obscures less spectacular activity: checking neighbors, seeking relatives, collecting supplies, and deciding what to do first.
Solnit's challenge does not rely solely on heartwarming anecdotes. Disaster sociology has a long history of field research. Summaries from the University of Delaware's Disaster Research Center repeatedly caution that panic and violence do not describe all post-disaster conduct; cooperation can exceed expectations. Even during evacuation, delay and reluctance to leave may require more attention than immediate uncontrolled flight.
The first chapter recounts Anna Holshouser's work after the 1906 earthquake. In Golden Gate Park she made a tent with blankets and other materials, sheltered people, and established a small soup kitchen. Initially lacking even utensils, the operation gradually received supplies and became known as Mizpah Café. Someone herself displaced was also caring for others.
Such activity changes the starting assumption. Expect robbery, and gatherings look suspicious; approach, and people may be sharing food, finding tools, and organizing temporary shelter. Needing outside help does not imply inability to judge immediate matters. Classifying everyone as victims awaiting placement may simplify counting without explaining their capacities.
Rescue therefore involves more than whether supplies arrive. Who can report actual needs, participate in arrangements, and preserve neighborhood ties also affects success. Professionals bring equipment and expertise; residents know particular people and places. Connecting these forms of knowledge can reduce avoidable mistakes.
Participation must not become a new condition, however. Someone can cook; someone else is seeking a missing relative; another is too exhausted to act. The last person still deserves care. Celebrating active survivors alone risks making usefulness a qualification for assistance.
That is also a caution about the title. Some find long-missed closeness in shared work. Others do not, and grief is not thereby negligible. Mutual aid matters because losses are real and people need one another, not because suffering has finally supplied meaning.
Attempts to Control Disorder Can Create It
Why is the capacity to cooperate so often missed? Solnit examines “elite panic”: people with power and resources may mistake losing their own control for society's collapse, then respond with excessive force.
San Francisco in 1906 presents a sharp example. Earthquake and fire devastated the city, while the mayor issued a proclamation claiming soldiers and police could shoot looters and other offenders. The National Park Service's account notes that he exceeded his authority. Armed groups acting separately produced contradictory orders and complaints about forced evacuations and shootings of suspects.
We need not pretend crime vanished. The question is why responding to real dangers expands into suspicion of an entire population. Once ordinary people are assumed to be potential rioters, taking supplies, returning home, or remaining outside can all become suspicious before anyone explains.
The situation then resists correction. Distrusting residents means hearing less of their judgment; hearing less makes them seem still less comprehensible. Those rescue should protect become objects against which rescue arrangements defend themselves. Solnit is strongest when showing these consequences rather than merely debating optimism or pessimism about human nature.
Yet she does not make all officials, police, and agencies identical. In a publication-time interview she explicitly credits conscientious police and Coast Guard rescuers. The National Park Service also records military firefighting, medical aid, communications, and supplies in 1906. Identifying coercive wrongs should not erase actual assistance.
Together these sides offer more than virtuous citizens opposed to heartless officials. Public institutions can mobilize resources and impose unchallengeable arrangements; local people know their surroundings but may lack equipment. Valuing mutual aid cannot remove public responsibility, just as indispensable expertise cannot require residents' silence.
Solnit's suspicion of institutions is strong. Readers may sometimes find the move from concrete rescue to social ideals too quick. Publishers Weekly's contemporary review registers that reservation: compelling historical accounts do not make the subsequent vision of a different everyday society equally well established.
The distinction matters. A few days of cooking and rescue cannot by themselves answer long-term housing, fiscal allocation, and professional maintenance. They can challenge the convenient assertion that only coercion makes people cooperate. Sustaining cooperation requires further work, not indefinitely extending an emergency into a model society.
Afterward, Who Can Still Speak?
Mexico City's garment workers give that afterward a more concrete form. During the 1985 earthquake, collapsed factories exposed a brutal ranking of workers' lives against employers' property. Discussing the history, Solnit describes employers attempting to salvage machinery while workers remained beneath rubble.
Mexico's education ministry also records survivors' further difficulties: lost jobs, unsettled compensation, and preexisting low pay and poor conditions. Rescue is the immediate priority. But if old relationships still decide what survivors may say or receive, surviving creates continuing difficulties.
The workers organized, registering a union in October of that year named for September 19. They were no longer only victims described by others. Unpaid wages, compensation, and future working conditions became collective demands.
Compared with a temporary street meal, this is slower and less buoyant. It involves organization, negotiation, and conflicting interests. Establishing a union does not guarantee every problem's solution. Yet asking what mutual support can leave behind requires attending precisely to less moving, harder work.
Two hopes emerge. One celebrates people stepping forward in emergencies, then hands everything back to normal. Another asks why people who have learned to act together should become passive again once the immediate danger eases. The garment workers' organization gives that question substance beyond nostalgia for solidarity.
The book thus differs from ordinary inspiration. It asks not merely for more individual kindness but which arrangements enable kindness to act and which dismantle existing cooperation. Readers need not accept Solnit's entire social vision to find it difficult afterward to describe disaster only through disorder, chaos, and management.
Paradise remains a risky word, potentially painful to someone bereaved. The useful inheritance need not be the label. It is the facts the label asks us to recognize: residents do more than wait; recipients of aid have knowledge and claims. After catastrophe, rebuilding concerns more than roads and homes. It also concerns whether people can continue participating in decisions about their lives.