For more than three decades, a chemical plant in Minamata, Japan, discharged wastewater carrying methylmercury into Minamata Bay and later into the wider Shiranui Sea. The mercury built up in the fish that coastal families ate every day. The poisoning that followed had no explosion, no release time, and no date anyone could circle. Years passed before the cause was officially named, and that delay held back protective action, regulation and compensation. This piece covers what the record establishes, why recognition took so long, why the victim count is still contested, and what a slow-onset event asks of agencies whose plans assume a sudden one.

The Chisso plant and the acetaldehyde process: what was discharged and for how long

Minamata is a small city on the southwest coast of Kumamoto Prefecture, on the island of Kyushu, facing the Shiranui Sea. The chemical plant there belonged to the company now called Chisso Corporation, which operated under earlier names before it took that name in 1965. The plant made fertilizer and a range of industrial chemicals, and the town’s economy depended on it heavily. In 1932 the plant began producing acetaldehyde, an intermediate used to make plasticizers and other products. The process used a mercury compound as a catalyst, and methylmercury formed as a byproduct inside the process. Wastewater carrying it went into Hyakken harbor at the edge of Minamata Bay. Chisso ended mercury-based acetaldehyde production at Minamata in May 1968, so the discharge period ran about 36 years.

Methylmercury did its damage through the food web, which is the mechanism planners need to understand. Plankton and small organisms take up methylmercury from the water. The fish and shellfish that eat them retain it, and each step up the chain concentrates it further. Fishing families around the bay ate fish and shellfish at nearly every meal, so they received far higher doses than anyone eating a mixed diet. Production at the plant rose through the 1950s along with Japan’s postwar chemical industry, and the dose in the bay rose with it. In 1958 Chisso redirected part of its effluent from Hyakken harbor to the mouth of the Minamata River. Patients then began to appear in communities farther along the Shiranui Sea coast that had not reported cases before, and the outlet was later moved back. The geographic spread of the disease therefore followed decisions made about a drain.

I was not involved in Minamata in any capacity, and I do not read Japanese source documents. For this piece I am working from English-language material published by Japan’s Ministry of the Environment and the National Institute for Minamata Disease, from the published clinical work of the Japanese physician Masazumi Harada, and from Timothy S. George’s history Minamata: Pollution and the Struggle for Democracy in Postwar Japan (Harvard, 2001). Japan treats a second outbreak caused by a different company’s plant on the Agano River in Niigata Prefecture, recognized in 1965, as a separate case. This piece mentions it only where it affects the Minamata timeline.

Cats, seabirds and fishing hamlets: the signals before and after May 1956

The environment showed the problem before any official body acknowledged it. Residents of the fishing hamlets around the bay reported fish floating on the surface and seabirds that could not fly properly. Cats, which ate fish scraps, convulsed, ran in circles and died, and some fell into the sea, which is why locals described the cats as dancing, and fishing catches declined over the same period. None of these observations reached a public health officer through any reporting system, because no system existed that would have routed a dead cat in a fishing village to one as evidence of a human hazard. Later investigation traced human cases back to the early 1950s, years before the official discovery date. Researchers have debated how much earlier unrecognized cases may have occurred, and I leave that question to the clinical literature.

The date Japan treats as the official discovery is May 1, 1956. On that day Hajime Hosokawa, director of the hospital Chisso ran for its own workers, reported to the Minamata public health center that he was seeing patients with an unknown disease of the central nervous system. His first patients included young children from a fishing family. The city, the prefecture and local physicians formed a committee to investigate, and they found more cases as soon as they looked. The fact that the first formal report came from the company’s own hospital became important later, because Hosokawa’s subsequent research put him in direct conflict with his employer.

The early response treated the disease as possibly infectious, which was a reasonable first hypothesis for a cluster of neurological cases in one community. Some patients were moved to isolation wards, and their homes were disinfected. That precaution produced lasting harm. Patients and their families came to be treated as carriers of contagion, and the stigma persisted after the infectious hypothesis was dropped. Accounts of Minamata describe families who hid symptoms and declined to seek certification for years afterward, in part because of what being identified had cost the first patients. The disease also split the town between families whose income came from the plant and families whose income came from the sea.

1956 to 1959: the university findings, the fish that stayed on sale and the counter-theories

A research group at the Kumamoto University medical school took up the investigation in 1956. By late that year it had concluded that the disease was a poisoning by a heavy metal acquired by eating fish and shellfish from the bay, and the factory’s wastewater was the obvious suspect. Identifying the specific agent took longer. The group examined several candidates, including manganese, selenium and thallium. In 1958 a visiting British neurologist, Douglas McAlpine, noted that the Minamata cases resembled the pattern of organic mercury poisoning described in British workers in 1940, which is known as the Hunter-Russell syndrome. In July 1959 the Kumamoto University group announced that organic mercury was the cause.

By 1957, protective action was already possible. The poisoning could be avoided by not eating fish from the bay, and the prefecture considered applying Japan’s Food Sanitation Act to prohibit the sale of fish caught there. According to the standard histories, the national Ministry of Health and Welfare replied that the act could not be applied, because no clear evidence showed that all the fish in the bay were toxic. Fishing was never legally banned. Fishing cooperatives imposed some voluntary restrictions, and those reduced but did not stop consumption, because fishing families had little else to eat or sell. The ministry treated proof that every fish was toxic as the threshold for protecting people from any of them, and that threshold kept the exposure going.

After the 1959 announcement, alternative explanations appeared. Some came from people associated with the chemical industry. They proposed that explosives dumped in the bay after the war were responsible, or that an amine compound from spoiled fish was the cause. Both theories were discredited, and later courts and the national government settled the cause as methylmercury from the Chisso plant. In November 1959 a subcommittee of the ministry’s food sanitation council reported that an organic mercury compound was the main cause. The subcommittee was dissolved soon afterward, and the question passed to other government bodies that did not reach a conclusion for years. Discredited theories still did their work, because as long as the official record called the cause unsettled, nobody was required to act.

1959 to 1968: the cat experiment, the treatment device and the sympathy money contract

Hosokawa continued his own experiments at the company hospital. In the most widely cited of them, a cat fed wastewater from the acetaldehyde process developed the disease in October 1959. According to the account that comes from his later testimony, the company told him to stop that line of work, and the result was not made public at the time. Hosokawa gave that testimony in 1970 for the patients’ lawsuit, close to his death. It became part of the evidence that Chisso had information pointing at its own process while it was publicly disputing the university’s findings. I report the experiment as his testimony describes it. I am not adding detail beyond what the histories I am relying on record.

Fishermen whose livelihoods had collapsed confronted the company in late 1959, and in November a protest at the factory turned violent. In December Chisso installed a wastewater treatment device called a Cyclator and held a public ceremony for it. The device was not designed to remove dissolved methylmercury, and the discharge continued. Because the ceremony told the public the problem was being handled, pressure fell at the moment it should have increased. On December 30, 1959, the company signed an agreement with a patients’ organization to pay what was called sympathy money. The amounts were small. The agreement included a clause stating that the patients would not seek further compensation even if the factory were later found to be the cause. In 1973 the Kumamoto District Court held that contract void as contrary to public order.

From 1960 the issue faded from national attention, and the plant kept producing acetaldehyde and discharging effluent. The Niigata outbreak, recognized in 1965 along the Agano River downstream of a Showa Denko plant using a similar process, brought the question back. On September 26, 1968, the national government officially declared that Minamata disease was caused by methylmercury compounds discharged from the Chisso plant. Chisso had ended the mercury process four months earlier, and the Showa Denko plant had stopped its process in 1965. By the late 1960s, petrochemical methods had displaced the mercury route in the industry. Nine years separated the university’s 1959 finding from the government’s declaration, and the discharge continued for nearly all of them.

What methylmercury does: severe cases, congenital cases and the ambiguous ones

Methylmercury damages the central nervous system. The classic signs in Minamata patients include loss of sensation in the hands and feet, narrowing of the visual field, loss of coordination (ataxia), hearing impairment, slurred speech and tremor. The severe cases of the 1950s progressed to convulsions, coma and death, sometimes within weeks or months of the first symptoms. Survivors of severe poisoning were often left with permanent disability. Physicians had no treatment that would reverse the nerve damage, so medicine could treat symptoms and provide care but could not undo the injury.

Congenital Minamata disease set the outbreak apart from most poisonings. Methylmercury crosses the placenta. Children born to women who ate contaminated fish during pregnancy were born with severe neurological disabilities that resembled cerebral palsy, and their mothers often had mild symptoms or none. Masazumi Harada, a physician who worked with Minamata patients for decades, documented these cases and argued that they had first been mistaken for ordinary cerebral palsy. For planners, this means the population most harmed may not be the population that was exposed, and an exposure registry that records only people with symptoms will miss the children of people without them.

The cases at the mild end of the range are the ones that drive the present-day dispute. Long-term, lower-dose exposure can produce sensory disturbance in the extremities without the full classic picture. Sensory loss in the hands and feet also occurs with diabetes, spinal disease and ordinary aging. A person who ate Shiranui Sea fish for twenty years and now, in old age, has numb hands may have methylmercury damage, an unrelated condition, or both, and no simple test settles the question decades after the exposure ended. Regulators and claimants did not disagree about whether methylmercury causes disease. They disagreed over where to draw a line through a large and ambiguous population.

The victim count: certified patients, settlement recipients and the fight over criteria

Formal recognition as a Minamata disease patient comes through certification. Under the national pollution health damage laws, the governors of Kumamoto and Kagoshima prefectures certify patients for the Minamata outbreak, acting on the advice of expert review committees. In 1977 the national Environment Agency issued criteria that, as applied, generally required a combination of symptoms rather than one symptom alone. Critics, including Harada, argued that those criteria excluded many people who had been genuinely harmed, particularly those with sensory disturbance as the main finding. Most applicants were rejected under those criteria.

The certified count published by Japan’s Ministry of the Environment has stayed in the low thousands. Over roughly the past two decades the ministry’s published figures for Kumamoto and Kagoshima have been a little above 2,200 certified patients, with several hundred more certified for the separate Niigata outbreak, and most certified patients have died. The ministry updates these figures, and anyone citing one should take the current number from the ministry and give the date it was published. A far larger number of people have received some form of payment or medical support without certification. The political settlement of 1995 provided lump sums to roughly ten thousand uncertified people with qualifying symptoms, which is the scale the standard histories give. The 2009 relief law drew applications from tens of thousands more before its application window closed in 2012. Advocates and some researchers argue that the population with methylmercury-related health damage is larger again, because large numbers of people around the Shiranui Sea ate local fish during the discharge years and no comprehensive health survey of that population was ever carried out.

These figures do not compete as estimates of a single quantity, so they should not be averaged or ranked. The certified figure counts people who met an administrative standard. The settlement and relief figures count people who met a different and looser standard in exchange for giving up legal claims. The advocates’ figure is an argument about the exposed population. Each answers a different question, and Japanese government bodies, patient organizations and courts have disputed which question matters. The dispute is political as well as medical, because the definition of a victim decides who pays and how much. I do not offer a single number for how many people Minamata disease harmed, and I am not aware of a source that can honestly provide one.

Courts and compensation: the 1973 ruling, state liability and the settlements

Certified patients filed suit against Chisso in 1969. On March 20, 1973, the Kumamoto District Court found the company negligent and voided the 1959 sympathy money contract. Later in 1973, patients reached a direct compensation agreement with Chisso that provided lump-sum payments and ongoing support to certified patients. The agreement tied compensation to certification, so from then on the certification criteria decided who was paid. The payments strained Chisso’s finances. Kumamoto Prefecture began issuing prefectural bonds in the late 1970s to keep the company able to pay. Under the 2009 relief law, Chisso transferred its operating business to a subsidiary, JNC, in 2011, and kept the compensation obligations in the parent company.

The national and prefectural governments’ own responsibility took much longer to establish. In 1995, under Prime Minister Tomiichi Murayama’s coalition government, a political settlement offered lump sums to uncertified claimants who withdrew their lawsuits, and most plaintiff groups accepted it. One group of patients who had moved to the Kansai region declined and continued its case. On October 15, 2004, Japan’s Supreme Court held the national government and Kumamoto Prefecture liable for failing to use their regulatory powers to stop the discharge after the end of 1959. That ruling placed a date on the government’s failure, and the date matches the period described above, when the cause was known and the drain was still open.

New claims followed the 2004 ruling, and Parliament passed the 2009 relief law, which accepted applications from 2010 until July 2012. In April 2013 the Supreme Court ruled that certification could not be refused merely because an applicant lacked the combination of symptoms the 1977 criteria described, and that each case required individual assessment. Litigation by people who were rejected or who missed the application window has continued into the 2020s. Lower courts reached opposite results in 2023 and 2024, with the Osaka District Court recognizing a group of plaintiffs and the Kumamoto District Court rejecting another group’s claims. In May 2024, Ministry of the Environment staff cut the microphones of patient representatives at a meeting with the environment minister, and the ministry later apologized. Readers who need the current state of the cases should follow the Ministry of the Environment’s published material and reporting on the individual suits. Seventy years after the discovery date, the process for deciding who is a victim is still in court.

Planning for an event with no onset: where incident-based doctrine falls short

Most American emergency management doctrine assumes an incident with a start. ICS organizes work into operational periods that begin when someone declares an incident. EOC activation levels are tied to events or forecasts. Stafford Act declarations specify an incident period, and the definitions that govern major disaster declarations were written mainly with sudden natural hazards in mind. In January 2016 President Obama approved an emergency declaration for the Flint, Michigan water crisis and denied the request for a major disaster declaration, because the contamination did not fit the statutory definition of a natural catastrophe. Readers should verify current declaration criteria with FEMA and their state emergency management agency, but the structural point holds. A slow contamination event can grow to mass casualty scale without passing any threshold the system is built to notice.

At Minamata the deciding moments were small administrative choices, each made under uncertainty: whether to route an animal death to a health officer, whether to apply a food safety law before every fish had been proven toxic, whether to treat a university finding as sufficient to act on while industry disputed it, and whether to accept a treatment device without verifying what it removed. A sudden-onset plan presents none of those questions, because the explosion or the flood answers them. A slow-onset event requires someone with authority to decide that a signal is serious enough to act on, and that role is usually held by an environmental health or public health official rather than by the emergency manager. The emergency manager’s practical job is to make sure that person exists, knows the EOC is available, and has a written path to request coordination before the cause is proven.

Minamata also shows that recovery depends on recognition. Compensation required proof of causation, and causation was officially established in 1968. Payment then depended on certification criteria that excluded most applicants, so for many people the delay in recognition turned into a delay of decades in support. Any slow-onset exposure in an American county will raise the same issues, including whose symptoms count, which records establish exposure, and who keeps those records. The site’s article on the exposure record at the World Trade Center pile covers the same problem from the responder side. An agency that writes down who was exposed, where, and when, at the time it happens, gives later claimants and investigators something better than memory to work from.

What to do at your agency

• Have your animal control supervisor and the county environmental health manager agree on a one-paragraph written procedure for reporting clusters of unexplained animal illness or death to the health department, and add it to the animal control standard operating procedures by the end of the month.
• Have your emergency manager write one paragraph into the existing EOP basic plan naming which public health or environmental health official may request EOC coordination for a slow-onset contamination event, and what that request requires, without waiting for a declared incident.
• Have your LEPC chair pull the EPA Toxics Release Inventory reports for facilities in the county and put a ten-minute review of any facility reporting releases of mercury or other persistent metals to water on the next scheduled LEPC agenda.
• Have your emergency manager or environmental health manager download your state’s current fish consumption guidelines, confirm whether each lake and river in the county is listed, and add the guideline link to the hazard section of the county mitigation plan at its next review.
• Have your public information officer draft a holding statement template for an environmental exposure of unconfirmed cause, with fill-in lines labeled “confirmed” and “not confirmed,” and file it with the other templates in the joint information center kit.
• Have your EMS medical director or QA lead ask the district health office’s epidemiologist, in writing, how EMS patient care data reaches syndromic surveillance and whom EMS should call about an unusual neurological cluster, and record the answer and the contact in the QA manual.

Takeaways

• The Chisso plant at Minamata discharged methylmercury from its acetaldehyde process from 1932 until May 1968, and the poisoning reached people through fish and shellfish rather than through any single release.
• Dying cats, failing seabirds and declining catches signaled the hazard before the official discovery date of May 1, 1956, and no reporting path carried those observations to public health officials.
• Kumamoto University identified organic mercury as the cause in July 1959, and the national government did not officially declare the Chisso plant the source until September 1968, after the discharge had ended.
• Industry-linked counter-theories blaming wartime explosives or spoiled fish were discredited, but they kept the official record unsettled long enough to delay regulation.
• The 1959 sympathy money contract barred further claims even if Chisso were later found responsible, and the Kumamoto District Court voided it in 1973.
• Japan’s Ministry of the Environment certifies a little over 2,200 patients for the Kumamoto and Kagoshima outbreak, far larger numbers received settlement or relief payments without certification, and patient groups and courts still dispute the criteria, so no single figure should be presented as the victim count.
• Japan’s Supreme Court held the national government and Kumamoto Prefecture liable in 2004 for failing to stop the discharge after the end of 1959.
• A slow-onset event crosses no threshold that incident-based plans watch for, so an agency needs a named official with authority to act on signals and a written path to the EOC before the cause is proven.