On 26 August, severe flooding struck the border region between Tibet in China and Nepal. A sudden collapse of glacier ice and rock at high altitude sent ice, stone and sediment into the river system, producing fast-moving floodwater and debris flows. Villages, roads, bridges and border facilities were destroyed within a very short time. Many residents, workers, tourists and pilgrims travelling towards Mount Kailash and Lake Manasarovar had no time to escape. By the evening of 27 August, an Associated Press tally had confirmed more than 390 deaths, with over 1,400 people still missing. Because rescuers had not yet fully reached some of the worst-affected areas, the final figures were likely to change.
United Nations Secretary-General António Guterres responded quickly. He expressed deep sadness over the flooding in China and Nepal, offered condolences to the families of those who had died, wished the injured a swift recovery and voiced solidarity with everyone affected. He also praised the rescuers working in dangerous conditions to save lives. UN agencies in Nepal then began coordinating with the government and humanitarian partners, while the World Health Organization, World Food Programme and UNICEF mobilised personnel, medical supplies, food and other assistance. The UN statement and relief arrangements at https://press.un.org/en/2026/sgsm23249.doc.htm showed that this was no longer being treated as a local disaster, but as a major humanitarian emergency receiving international attention.
Several countries also acted quickly. India sent tents, medicines, blankets and hygiene supplies to Nepal. China coordinated emergency supplies and financial assistance. South Korea dispatched responders and pledged humanitarian aid, while the United Kingdom, the United States, Australia and the European Union announced funding, personnel or disaster-relief resources. The International Federation of Red Cross and Red Crescent Societies, the World Health Organization, the World Bank and Médecins Sans Frontières also became involved. This attention and assistance were entirely necessary. When hundreds of people die suddenly, more than a thousand remain missing and many families are destroyed within hours, no government or public institution with even a basic humanitarian commitment can remain indifferent.
Yet this rapid concentration of global attention made me think about another set of numbers that now rarely shocks us.
The war in Ukraine has continued for more than four years. Wartime deaths are difficult to count accurately. Both sides have reasons to conceal, minimise or inflate their losses, so no estimate should be treated as exact. Even so, long-term figures can reveal the approximate scale. The Center for Strategic and International Studies estimated that between February 2022 and December 2025, about 275,000 to 325,000 Russian service personnel and 100,000 to 140,000 Ukrainian service personnel were killed. Together, that amounts to roughly 375,000 to 465,000 military deaths, or an average of about 267 to 331 deaths every day. Over the same period, combined military casualties — killed, wounded and missing — were estimated at 1.7 to 1.8 million, averaging more than 1,200 a day. The CSIS estimate at https://www.csis.org/analysis/russias-grinding-war-ukraine cannot tell us exactly how many people died on any particular day, but it is enough to show the scale of loss the war has sustained over time.
The civilian figures are also grave. The UN Human Rights Monitoring Mission verified that in July 2026 alone, at least 437 civilians were killed and 2,610 injured in Ukraine. That is an average of around 14 civilian deaths and 84 injuries every day. These are only the cases verified individually. The UN also noted that access to some occupied territories and places where the early fighting was most intense remained limited, making the real toll likely to be considerably higher. The UN civilian casualty report for July 2026 at https://ukraine.ohchr.org/sites/default/files/2026-08/Ukraine%20-%20protection%20of%20civilians%20in%20armed%20conflict%20%28July%29_ENG.pdf does not include the military losses of either side.
Once these figures are placed alongside each other, an uncomfortable comparison emerges. The number of deaths so far confirmed in the China–Nepal floods is roughly equivalent to one to one and a half days of average military deaths in the Ukraine war. The flood deaths were concentrated within minutes or hours. Death on a comparable scale in the war has continued day after day for years.
The purpose of this comparison is not to diminish the seriousness of a natural disaster, nor to rank one kind of death against another. Behind every life that suddenly disappears is a family, along with relationships and a way of life that cannot be restored. The comparison raises a different question. Why can a natural disaster that kills hundreds of people rapidly shake the world, while human-caused deaths occurring every day on a similar or greater scale gradually recede into the background of the news?
Morally, deliberate killing by human beings should be even harder to tolerate than death caused by nature. Natural disasters may be connected to human-driven climate change, infrastructure decisions and failures of risk management, but they are not, in themselves, organised attempts by one group of people to kill another with weapons. War is decided by people, carried out by people and, at least in principle, can be stopped by people. Yet our reactions often move in the opposite direction. We are shocked and saddened by sudden disaster, and we follow the rising death toll. Faced with a prolonged war, however, we easily come to treat daily losses as an existing condition.
This is not simply because people have lost their capacity for compassion. The deeper problem is that human attention is not allocated according to the total amount of suffering. Attention first detects change rather than conditions that persist. A sudden flood has a clear starting point, a concentrated location, powerful images, a changing number of missing people and an unfolding rescue operation. It abruptly breaks the previous order of life and therefore creates a strong difference. War is dispersed across a long period, a wide area and contested sets of statistics. Today’s shelling resembles yesterday’s. Today’s death toll resembles last month’s. When similar information is repeated, the human response gradually weakens.
Psychology calls this process habituation. It is one of our basic capacities for survival. If we responded to every continuing stimulus with its original intensity, ordinary life would become impossible. Yet the same capacity can have serious moral consequences. When death keeps recurring, it is no longer perceived as a new event. It is absorbed into the environment that is already there.
Psychological research has also described “psychic numbing” and “compassion fade”. A victim with a name, a face and a recognisable life often evokes strong sympathy. Once victims become tens, hundreds or hundreds of thousands, our emotional response does not grow with the numbers and may even decline. Some research suggests that compassion can begin to weaken when the number of victims rises from one to two. The experimental work on compassion fade at https://pmc.ncbi.nlm.nih.gov/articles/PMC4062481/ shows that people can understand numbers intellectually without being able to hold, at an emotional level, all the lives those numbers represent.
Psychology alone, however, is not enough to explain what happens. Our attention is shaped not only by the limits of cognition, but also by the combined activity of media organisations, digital platforms, governments and public institutions.
Commercial media need new events. A sudden flood has a clear beginning, dramatic images and a rescue process that can be updated continuously, making it easy to form a complete news cycle. A prolonged war has no comparable boundary. Unless there is a major attack, the fall of an important city or a sudden political shift, routine deaths struggle to become daily headlines.
Platform algorithms follow a similar logic. They do not directly judge which suffering is more serious. They infer what should continue to circulate from clicks, viewing time, sharing, comments and emotional reactions. A sudden disaster produces concentrated engagement and therefore receives more recommendations. When similar reports from a long war continue to appear and users respond less strongly, the platforms reduce their visibility. Algorithms do not declare that some lives matter less, but by deciding what can be seen, they take part in the public ordering of whose lives appear important.
Governments and international institutions operate under another set of constraints. Condolences after a natural disaster usually create little political controversy and do not require the speaker to assign responsibility immediately. Countries can mourn the victims together, praise rescuers and provide material assistance. This is a comparatively safe and unambiguous humanitarian position. War is different. To discuss deaths in war is often to confront further questions about responsibility, ceasefires, arms supplies, sanctions and the costs one’s own country is prepared to bear. Once compassion enters the context of war, it is quickly absorbed into national positions, alliances and strategic objectives.
Lives lost in war are therefore readily translated into another language. People are no longer described as particular human beings, but as manpower, losses, targets, gains, exchange ratios or changes along the front line. Death has not disappeared. It has been recoded in military and political terms. A person’s place within a family, and their relationships with parents, partners and children, are compressed into a single entry in a statistical table.
This is what I mean by the manipulation of attention. Manipulation here does not necessarily imply a single hidden operator. Some forms are clearly deliberate. Propaganda agencies choose what to disclose, media outlets construct headlines, platforms design recommendation systems and governments control casualty information. Yet much of the process is not planned by any one person. Commercial organisations seek efficient distribution, governments manage political risk, international bodies observe diplomatic boundaries, platforms pursue user response and ordinary people protect their own psychological capacity. Each participant follows a local set of rules, but together they produce a stable order of attention.
That order determines what suddenly enters our field of vision and what remains in the background. It also determines which deaths become public events and which are merely recorded. We imagine that we are looking at the world freely. In reality, we see only the part of the world that remains visible after being filtered through commercial, political, technological and psychological mechanisms.
From the perspective of Sustenesis, this process can be understood as an outcome jointly formed by Difference, Constraint and Sustained Coherence. A sudden disaster creates an obvious difference and therefore enters attention quickly. Repeated deaths in war gradually form a stable background. That stability is not reasonable or acceptable. It means only that war has become a continuing structure within the media environment. Commercial interests, political positions, institutional rules, algorithmic systems and the limits of human emotional endurance create multiple constraints. Together, they sustain a relatively stable order of visibility in which some events are repeatedly amplified while more persistent forms of suffering are pushed down.
The real danger is not that we have lost compassion altogether. It is that repetition can make the most intolerable conditions appear normal. Three hundred deaths on the first day would shake the world. If three hundred people die every day, then after several months “three hundred a day” may no longer count as news. The deaths have not diminished. What has changed is their position within the structure of attention.
We cannot expect everyone to live each day with the emotional intensity they felt when they first heard news of the war. That would be neither realistic nor psychologically bearable. Society can, however, build mechanisms that resist numbness. Media organisations can regularly present cumulative deaths and average daily losses, rather than reporting only territorial changes along the front. Public institutions can continue distinguishing between those killed, injured and missing, instead of allowing the broad word “casualties” to obscure actual consequences. Reporting can reconnect numbers with particular people, families and lives, repeatedly returning a prolonged war to a human scale.
Natural disasters deserve the world’s attention. The problem is not that we have given too much attention to this flood, but that we give too little to human-caused deaths that continue over time. We cannot stop every force of nature. War, at least in principle, can be stopped by people. A world capable of mourning hundreds of people killed in a sudden disaster should not, at the same time, treat the daily killing of hundreds of people as a normal background condition.
Human beings do not allocate attention according to the severity of suffering. We allocate it according to suffering’s ability to create a difference within systems of communication. Once a form of death no longer creates a new difference, it gradually leaves our field of perception. War’s deepest victory may not be the amount of territory it captures, but the moment when continuing deaths no longer surprise us.
Discover more from Geoffrey Chen
Subscribe to get the latest posts sent to your email.